RO.T.DEMAHDLOK.mao.R
`
`NUMBERZ 3? 1988.
`
`IMMUNOGEN 2304, pg. 1
`Ph
`igenlx v
`. Immunogen
`IPR2014-00676
`
`IMMUNOGEN 2304, pg. 1
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`MOLECULAR
`BIOTHERAPY
`
`THE INTERNATIONAL JOURNAL FOR THE APPLICATION
`OF BIOLOGICALS IN CLINICAL OR VETER[NARY PRACTICE
`
`VOLUME l • NUMBER 2 • 1988
`
`ROBERT K. OLDHAM, EDITOR
`
`58 Editorial
`
`Review
`
`61 An integration of cultivation and purification in membrane
`bioreactors: production of monoclonal antibodies and
`lymphokines by microencapsulated hybridomas
`A. Prokop
`
`Papers
`
`74 Host immune responses to tumor cells augmented by
`bleomycin and their therapeutic effects on rat fibrosarcoma
`M. Hosokawa, Z-y Xu, K. Morikawa, J-i Hamada, and
`H. Kobayashi
`
`81
`
`Toxicities associated with monoclonal antibody infusions in
`cancer patients
`R. 0. Dillman, J. C. Beauregard, M. Jamieson, D. Amox, and
`S. E. Halpern
`
`86 Report of two cases of acute myelogenous leukemia
`immunized with autologous leukemia-derived hybrid cells
`E. P. Cohen, V. A. Lazda, S. G. Schade, J. L. Kennedy,
`E. R. Kaufman, and K. L. Hagen
`
`· 96 Mechanisms of TNF resistance: identification of membrane
`phosphoproteins associated with a dominant resistant
`phenotype in lymphoid-myeloid somatic cell hybrids
`S. Schiitze, S. Nottrott, P. Scheurich, M. Kronke, and
`K. Pfizenmaier
`
`103 Adriamycin custom-tailored immunoconjugates in the
`treatment of human malignancies
`R. K. Oldham, M. Lewis, D. W. Orr, B. Avner, S-K Liao,
`J. R. Ogden, B. Avner, and R. Birch
`
`114 Conference Report
`
`116 Patent Reports
`
`118 Book Review
`
`119 Calendar
`
`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,
`MA 02180.
`
`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:
`(617) 744-3350). quoting the following:
`0952-8172/88$3.00 + 0.
`© 1988 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., 1988, vol. 1, no. 2 57
`
`IMMUNOGEN 2304, pg. 2
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Adriamycin custom-tailored immunoconjugates in
`the treatment of human malignancies
`
`Robert K. Oldham, MD, Marvin Lewis, MD, Douglas W. Orr, MD,
`Barry Avner, PhD, Shuen-Kuei Liao, PhD, John R. Ogden, PhD, Belina Avner, BA,
`and Robert Birch, PhD
`Williamson Medical Center, Biological Therapy Institute, and Biotherapeutics, Inc., Franklin, TN
`
`Twenty-three patients with disseminated refractory malignancies each received a tailored combination
`of adriamycin-conjugated murine monoclonal antibodies. Tumors were typed using a panel of antibod(cid:173)
`ies. Cocktails of up to six antibodies were selected based on binding greater than 80% of the malignant
`cells as tested by immunoperoxidase and flow cytometry. These monoclonal antibodies were then
`conjugated to Adriamycin and administered intravenously. Seventeen of23 patients had reactions to
`the administration ofimmunoconjugates, but these were tolerable in all but two patients. Fever, chills,
`pruritis, and skin rash were by far the most common transitory reactions. All were well controlled with
`premedication. In several patients there was limited antigenic drift among various biopsies within the
`same patient over time. This observation confirms the necessity for the use of a cocktail of antibodies
`if one wishes to cover all tumor cells. Preliminary serologic evidence suggests that the development of
`an !gM antibody, which is specific against the mouse monoclonal antibody, has the specificity and
`sensitivity to predict clinical reactions. Selected patients were re-treated. One patient with chronic
`lymphocytic leukemia had re-treatment on three occasions and demonstrated regression of peripheral
`lymph nodes. Two patients with breast carcinoma had definite improvement in ulcerating skin lesions
`and two patients with tongue carcinoma had shrinkage of their lesions. In the- course of the study free
`Adriamycin released from the monoclonal antibodies was discovered to be a limiting factor in the
`amount of antibody that could be administered. Up to I g of Adriamycin and up to 5 g of monoclonal
`antibody were administered. The limiting factor appeared to be a variable dissociation of active
`Adriamycin from the antibody that unpredictably caused hemopoietic depression. This study demon(cid:173)
`strates the feasibility and reviews technical considerations in preparing immunoconjugate cocktails for
`patients with refractory malignancies. The major technical hurdle appears to be the selection of an
`effective conjugation method that can be used to optimally bind Adriamycin to monoclonal antibodies
`-
`for targeted cancer therapy.
`
`Keywords: Adriamycin; immunoconjugate cocktails; targeted cancer therapy
`
`Introduction
`Since Kohler and Milstein 1 provided the technique by
`which monoclonal antibodies could be produced in
`virtually unlimited quantities, there has been an ex(cid:173)
`plosion in the use of monoclonal antibodies in patients
`with malignancies. This paper introduces the concept
`of combination monoclonal antibodies, specifically
`tailored for individual patients, combined with Adria(cid:173)
`mycin.
`The hypothesis that a combination of monoclonal
`antibodies would be necessary to cover virtually all
`cancer cells in a variety of sites and that each patient
`
`Address reprint requests to Dr. Robert K. Oldham, Biological Ther(cid:173)
`apy Institute, Hospital Drive, PO Box 1676, Franklin, TN 37065-
`1676, USA
`Accepted for publication September 16, 1988.
`
`would require an individually specified immunoconju(cid:173)
`gate dominated in this research. Single monoclonal
`antibodies have been demonstrated with immunoper(cid:173)
`oxidase tissue stains and radioisotopes to localize in
`areas of malignancy and to individual malignant
`cells. 2•3 However, it is well known that cancer cells
`have a variety of antigens, which are not cancer-spe(cid:173)
`cific and which can vary within clusters of tumor cells
`both in one location and in distant metastatic sites
`(microheterogeneity). Tumor antigens may also vary
`during phases of tumor cell maturation. In addition,
`we have typed over 100 tumors from different pa(cid:173)
`tients, and quantitative differences are frequent. No
`two have demonstrated the same typing pattern (ma(cid:173)
`croheterogeneity). Heterogeneity is basic to the thesis
`explored here. Thus, an attempt was made to identify
`a combination of antibodies, which could potentially
`recognize up to 100% of malignant cells within a vari(cid:173)
`ety of primary and metastatic sites. This was done by
`
`Mol. Biother., 1988, vol. 1, no. 2
`
`© 1988 Butterworth Publishers 103
`
`IMMUNOGEN 2304, pg. 3
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Adriamycin immunoconjugates: R.K. Oldham et at.
`
`making a large number of monoclonal antibodies
`against tumors and then typing individual patient's
`tumor biopsies and demonstrating attachment to
`greater than 80% of the cells within the malignancy. To
`that end, cocktails of as many as six antibodies were
`administered to patients following conjugation with
`Adriamycin.
`These antibodies were more than 95% pure, main-
`
`tained immunoreactivity after conjugation, and were
`tested for safety in a variety of systems prior to ad(cid:173)
`ministration to patients (Figure 1). This paper demon(cid:173)
`strates the feasibility of treating patients with mixtures
`of monoclonal immunoconjugates and addresses tech(cid:173)
`nical considerations involved in the process. Observa(cid:173)
`tions of side effects, the re-treatment of patients sub(cid:173)
`sequently with similar or identical antibodies, and the
`
`Product Testing and Analysis
`
`Nude Mouse Test
`
`Xenoeratt
`
`Storage
`lhfrlgerole
`
`Biochemical
`Analysis
`Protein
`
`DIPLC
`
`Immunological
`Analysis
`Hllllology
`
`PCFIA
`FACS
`
`/
`
`Safety Test
`
`~{}···
`. • -
`J
`~
`
`<
`
`Sterility Assay
`___ _.,...,~ Tr,-ptlcaaa Soy lllroth
`
`Thloelycollate Broth
`
`Saaourautl'e Aear
`
`~ Endotoxin Assay
`
`&.lllluloUI Alllloeaacyte &.yeah (LA&.!
`Chromogenic Aaaay
`
`In vitro Cytotoxicity Assay
`
`Taroat cell
`Non-target cell
`Figure 1. After preparation of immunoconjugate and prior to use in humans, extensive testing of immunoconjugate was done and is
`described elsewhere.8 Abbreviations: SDS·PAGE = sodium dodecyl sulfate polyacrilamide gel electrophoresis; HPLC = high perfor(cid:173)
`mance liquid chromatography; FACS = fluorescent-activated cell sorter; PCFIA = Particle concentration fluorescence immunoassay.
`
`104
`
`Mol. Biother., 1988, vol. 1, no. 2
`
`IMMUNOGEN 2304, pg. 4
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Papers
`
`Table 1. Monoclonal antibodies used in the present
`study
`
`MoAb
`
`lsotype
`
`Immunogen
`
`Antigen
`structure
`
`ND•
`
`ND
`
`>300kD glyco-
`protein
`
`ND
`
`220kD-400kD
`glycoprotein
`ND
`
`29kD + 31kD
`protein
`
`ND
`
`ND
`
`ND
`
`29kD + 31kD
`
`ND
`
`ND
`
`95kD-150kD
`glycoprotein
`p97-like (97kD)
`glycoprotein
`110kD protein
`110kD + 40kD
`protein
`ND
`
`view Board of that institution. Patients were referred
`primarily by oncologists after failure of standard mo(cid:173)
`dalities of treatment. Each patient was initially seen
`by a medical oncologist who reviewed the history and
`medical records, confirmed the lack of standard thera(cid:173)
`peutic options available to the patient, informed the
`patient of the experimental nature of the study, and
`had a full discussion with each patient of the strategy
`involved in this therapy and of other experimental
`therapeutic options available. After a determination of
`suitability for the study and informed consent, tissue
`samples were obtained by biopsy. All typing was done
`on frozen tissue, either directly or on tissue that had
`been expanded by a xenograft in nude mice or by tis(cid:173)
`sue culture propagation. Antibody selection was by
`immunoperoxidase and flow cytometry as described
`in detail elsewhere.4·5 A minimal period of 45 days
`was necessary for tissue typing and preparation of
`sufficient quantities of immunoconjugate for treat(cid:173)
`ment. A typical regimen consists of 3 days for typing
`the tissue with a panel of monoclonal antibodies, ape(cid:173)
`riod of 4 weeks for production of sufficient quantities
`of the individual antibodies followed by conjugation of
`Adriamycin, and extensive safety testing over a final 3
`weeks (Figure 1). Thus, within 3 months patients were
`seen and treated with a tailored combination of anti(cid:173)
`bodies conjugated to Adriamycin.
`Immunoconjugate preparations dissolved in normal
`saline were given on a Monday, Wednesday, and Fri(cid:173)
`day over a period of 1 to 5 hours. Total amounts of
`immunoconjugate were generally given over a 2- to 3-
`week period. An initial test dose of 10 mg of Adriamy(cid:173)
`cin bound to antibody was given. The dose was then
`quickly escalated depending on the phase of the study.
`Early in the investigation total Adriamycin doses were
`kept below 300 mg. Near the end of the investigation,
`antibody amounts were escalated to try to give as
`much as 1 g of Adriamycin and 3 to 5 g of antibody
`over a period of 2 to 3 weeks. A registered nurse was
`always available during administration, and patients
`were premedicated with acetaminophen and diphen(cid:173)
`hydramine for fevers, meperidine for rigors, and epi(cid:173)
`nephrine in four patients for significant allergic reac(cid:173)
`tions.
`
`Antibody selection and preparation
`Immunization of mice and preparation of hybridomas
`are described elswewhere. 4·5 Over 100 antibodies
`were available for tissue typing, and we selected 28 for
`the standard panel. Seven of these were acquired else(cid:173)
`where and 21 were produced in the laboratories of
`Biotherapeutics. Five of these originated from immu(cid:173)
`nization with breast cancers, 11 from melanomas, 3
`from adenocarcinomas of the kidney, 2 from an islet
`cell carcinoma of the pancreas, and 7 from colon car(cid:173)
`cinomas.4 The majority of the antibodies were IgG1
`with the exception of two lgG2's (melanoma) and five
`lgG3's (colon carcinomas). Table 1 illustrates the
`characteristics of 18 antibodies from the panel used in
`this clinical study.
`
`BA-Br-1
`
`BA-Br-2
`
`BA-Br-3
`
`BT-Br-4
`
`BA-Br-5
`
`BT-Br-6
`
`BT-Co-1b
`
`BT-Co-2
`
`BT-Co-3
`
`BT-Co-4
`
`BT-Co-5b
`
`BT-Co-6
`
`BT-Me-3
`
`BT-Me-4
`
`BT-Me-5
`
`BT-Me-7
`BT-Me-8
`
`lgG1
`
`lgG1
`
`lgG1
`
`lgG1
`
`lgG1
`
`lgG1
`
`lgG3
`
`lgG3
`
`lgG3
`
`lgG3
`
`lgG3
`
`Membrane extract of
`breast carcinoma tissue
`Dispersed cells from
`breast carcinoma tissue
`Membrane extract of
`breast carcinoma cell
`line CAMA-1
`Dispersed cells from
`breast carcinoma tissue
`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
`Dispersed cells from
`melanoma tissue
`Melanoma cell line CaCL
`78-1
`lgG2a Melanoma cell line CaCL
`78-1
`Melanoma cell line BUR
`Melanoma cell line BUR
`
`lgG1
`
`lgG1
`
`lgG1
`
`lgG1
`lgG1
`
`BT-Ne-3
`
`lgG1
`
`Dispersed cells from
`hypernephroma tissue
`
`All antibodies are referenced to Liao et al. 5
`• Not yet defined, although attempts were made to determine the molec(cid:173)
`ular mass of antigen involved.
`b Based on epitope blocking and indirect immunoprecipitation experi(cid:173)
`ments, BT-Co-1 and BT-Co-5 recognized different epitopes residing on
`the same or similar molecules.
`
`biologic effects and tumor localization of the antibod(cid:173)
`ies, as well as the efficacy of this treatent, are pre(cid:173)
`sented.
`
`Materials and Methods
`Patient selection
`This clinical trial was carried out in Williamson Medi(cid:173)
`cal Center after approval by the Investigational Re-
`
`Mol. Biother., 1988; val. 1, no. 2
`
`105
`
`IMMUNOGEN 2304, pg. 5
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Table 2. Disease categories
`
`Breast carcinoma
`CLL
`Ovarian
`Tongue
`Renal
`Rectal
`Prostate
`Cholangiocarcinoma
`Sarcoma
`Parotid
`Lung
`Total
`
`10
`1
`1
`2
`1
`1
`1
`1
`1
`1
`2
`
`23
`
`The initial selection of antibodies was done by
`immunohistochemical phenotyping6-10 and is fully
`described elsewhere. 5 Antibodies were selected
`immunohistochemically by immunoperoxidase. The
`selection of antibodies was based on a grading system
`of 1 to 4 + , which included a judgment both on the
`intensity of staining as well as the distribution of
`staining and the specific characteristics of the stain(cid:173)
`ing. The variation in grading between observers was
`less than 10%. The selection by the same observers
`was reproducible over 90% of the time. Staining pat(cid:173)
`terns varied from homogeneous staining of membrane
`and/or cytoplasm to patchwork staining of given
`tumor areas leaving adjacent tumor areas virtually un(cid:173)
`stained, to scattered reactivity of tumor cells in a
`"sea" of nonreactive tumor cells. 5 Selections of anti(cid:173)
`bodies were made and encompassed considerations of
`intensity, distribution, and patterns of staining. Posi(cid:173)
`tive controls included anti-HLA; negative controls in(cid:173)
`cluded nonspecific random mouse immunoglobulin in
`all sections. All results were scored independently by
`two scientists.
`Based on these considerations, a combination of
`antibodies was selected, which was thought to cover
`more than 80% of tumor cells in the specimen. Where
`possible, these were subsequently checked by flow
`cytometry. This technique used single cell suspen(cid:173)
`sions of patients' tumors and quantitated the percent
`of viable cells bound by antibody, as well as the inten-
`
`Table 3. Selected characteristics of patients
`
`Adriamycin immunoconjugates: R.K. Oldham eta!.
`
`sity of binding, to ensure that combinations of anti(cid:173)
`bodies would be additive or synergistic and would not
`interfere with binding of each other. In flow cytometry
`binding is quantitated by measuring the percentage of
`tumor cells that bind the antibody (percent positive)
`and by the mean peak channel (MPC), which signifies
`the intensity of loading of the antibody on each cell
`(i.e., how many molecules of antibody bind to each
`positive cell.) Of the 23 patients included in the sub(cid:173)
`ject study, flow cytometry analysis was performed on
`tumor biopsy cells from eight of the patients (35%).
`The agreement between positives on immunohisto(cid:173)
`chemistry and positives on flow cytometry was 86%.
`Interaction of different antibodies was determined on
`flow cytometry by measurement of the above-de(cid:173)
`scribed parameters on cells exposed to two or more of
`the antibodies simultaneously, or sequentially. Com(cid:173)
`parison of the percent positive and PMC obtained in
`these mixing studies, with those obtained for the sin(cid:173)
`gle antibody, enabled us to determine the degree to
`which the antibodies interacted, either additively or
`subtractively. These studies demonstrate that greater
`than 80% of the cells were able to be coated with anti(cid:173)
`bodies.
`Mter selection of appropriate monoclonal antibod(cid:173)
`ies, production was scaled up in ascites, and these an(cid:173)
`tibodies were purified in gram amounts and chemical(cid:173)
`ly conjugated with Adriamycin. 11 Immunoconjugates
`of Adriamycin and murine antibody using a cis-aconi(cid:173)
`tate linker were prepared by a modification of the
`method of Shen and Ryser. 12•13 Clinical grade Adria(cid:173)
`mycin containing lactose (Adria Labs) was converted
`to a carboxylic acid intermediate by reacting cis-acon(cid:173)
`itic anhydride with the free Adriamycin base. This
`compound was in turn converted to the active ester
`by its reaction with 1-ethyl-3-(3-dimethylamino-pro(cid:173)
`pyl) carbodiimide (EDC) and N-hydroxysuccinimide
`(NHS). The activated Adriamycin solution was mixed
`with purified antibody and stirred overnight at room
`temperature in the dark. Free Adriamycin was re(cid:173)
`moved by tangential flow ultrafiltration using MWCO
`30 000 membranes. Adriamycin-to-antibody molar
`ratios were calculated from protein concentrations es(cid:173)
`timated by the Bradford Coomassie Blue binding
`
`Performance status
`
`Age
`
`Sex
`
`Previous chemotherapy
`Previous Adriamycin
`Site of disease
`
`Source biopsy
`
`>70%
`10
`
`>50 yr
`12
`Male
`11
`
`Soft tissue
`only
`1
`
`Viscera
`7
`
`19/23
`12/23
`
`Effusion
`2
`
`<70%
`13
`
`<50 yr
`11
`Female
`12
`
`visceral
`+lor bone
`22
`
`Primary
`2
`
`Skin subcutaneous
`3
`
`Lymph
`node
`9
`
`106
`
`Mol. Biother., 1988, val. 1, no. 2
`
`IMMUNOGEN 2304, pg. 6
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Papers
`
`Table 4. Immunohistochemical results of metastatic melanoma lesions removed at various times from one patient
`(70.23)
`
`Occipital
`LN met
`01/23/86
`
`Tissue culture
`cells from
`01/23/86 LN met
`
`Mediastinal
`LN met
`05/05/86
`
`Neck
`LN met
`04/27/87
`
`Supraclavicular
`LN met
`06/15/87
`
`Femoral
`LN met
`06/15/87
`
`Brain
`met
`09/28/87
`
`4+
`4+
`
`2+
`
`4+
`4+
`3+
`
`4+
`4+
`
`1+
`
`4+
`4+
`3+
`
`4+
`4+
`
`4+
`1+
`
`1+
`3+
`
`4+
`4+
`4+
`
`4+
`4+
`
`4+
`2+
`
`1+
`3+
`
`4+
`4+
`4+
`
`4+
`4+
`
`4+
`2+
`
`2+
`2+
`
`2+
`1+
`NE
`
`Anti melanoma
`BT-Me-7
`BT-Me-8
`BT-Me-3
`BT-Me-4
`BT-Me-5
`Antibreast CA
`BA-Br-1
`BA-Br-3
`BT-Br-4
`BA-Br-5
`BA-Br-6
`R-11
`R-13
`Anticolon CA
`CO-Co-1
`BT-Co-2
`BT-Co-3
`BT-Co-4
`BT-Co-5
`BR-Co-6
`BR-Co-7
`BR-Co-8
`BR-Co-9
`Antirenal CA
`BT-Ne-3
`
`4+
`4+
`4+
`4+
`
`3+
`
`4+
`
`1+
`
`1+
`2+
`4+
`4+
`
`1+
`
`NT
`NT
`NT
`NT
`
`1+
`
`Abbreviations: NE = not available because tissue section was washed off; NT = not tested; CA = carcinoma.
`
`assay and Adriamycin concentrations estimated from
`a standard curve of absorbance at 495 nm. Gel perme(cid:173)
`ation HPLC with a Superose 6 column (Pharmacia)
`and SDS-PAGE were performed to examine the ho(cid:173)
`mogeneity of antibody. Immunoconjugates were ster(cid:173)
`ilized by filtration and stored in the dark at 4°C until
`used.
`Adriamycin was tightly associated with antibody,
`but the exact nature of the linkage is unknown. 11 We
`believe some Adriamycin was covalently linked but a
`majority was noncovalently associated with antibody.
`Association of Adriamycin molecules to antibody was
`stable for at least 6 months in phosphate-buffered sa(cid:173)
`line and would tolerate gel filtration under mild condi(cid:173)
`tions; stringent competition for binding or denatura(cid:173)
`tion (incubation with Amberlite XAD-2 or SDS-PAGE
`respectively), however, revealed some Adriamycin
`that was not covalently linked to antibody. General
`safety tests revealed that nonspecific toxicities could
`occur if immunoconjugate levels were high enough.
`Early preparations using these methods demonstrated
`little clinical toxicity while later preparations proved
`less stable in serum and caused mild to moderate
`Adriamycin toxicity. Concurrent with this change in
`toxicity and stability was a formulation change by
`Adria Labs involving the addition of methylparaben.
`While not encumbering the chemical reactions in(cid:173)
`volved .in the conjugation, methylparaben may affect
`the strength of the noncovalent association between
`antibody and Adriamycin and thus decrease the stabil-
`
`Mol. Biother., 1988; vol. 1, no. 2
`
`ity of the complex in serum and lead to the release of
`more Adriamycin.
`Analysis by flow cytometry indicated that these
`immunoconjugates retained immunoreactivity after
`conjugation. Conjugates prepared in this manner ex(cid:173)
`hibited antibody specific in vitro cell-killing proper(cid:173)
`ties. Animal studies indicated a 15-fold lower nonspe(cid:173)
`cific lethal toxicity from the conjugate when compared
`with free Adriamycin. Examination of the bone mar(cid:173)
`row of treated animals also revealed a decreased tox(cid:173)
`icity associated with the conjugates. The conjugates
`also failed to produce soft tissue necrosis when inject(cid:173)
`ed intradermally at levels of Adriamycin, which in an
`unconjugated state cause severe damage. 11
`
`Clinical monitoring
`Testing of serum samples for human anti-mouse immu(cid:173)
`noglobulin was done with a fluorescent immunoassay
`using particle concentration. 14•15 Multiple samples
`were tested to detect the development and effect of
`antiglobulins. 15 Patient biopsies were tested for the
`presence of murine antibody by immunoperoxidase
`histochemistry or by flow cytometry using goat anti(cid:173)
`5
`mouse antibodies as the developing agent. 4
`•
`
`Results
`Tables 2 and 3 summarize the patient characteristics.
`For this report, 23 patients treated with immunoconju-
`
`107
`
`IMMUNOGEN 2304, pg. 7
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Table 5.
`Immunohistological typing of metastatic le(cid:173)
`sions at different times
`
`Table 6. Results of immunohistochemical tumor typing
`of two lymph node metastasis with our panel of mono(cid:173)
`clonal antibodies
`
`Adriamycin immunoconjugates: R.K. Oldham eta!.
`
`Tissue culture
`cells from
`left groin
`lymph node
`2/27/85
`
`Melanoma
`metastasis
`from
`right lung
`12/22/86
`
`Melanoma
`metastasis
`from
`left lung
`7/23/87
`
`MoAb
`
`Tested 1 0/31 /86
`
`Tested 11/12/87
`
`Reactivity
`
`MoAb
`developed
`from
`
`Breast
`BA-Br-1
`BA-Br-2
`BA-Br-3
`BT-Br-4
`BA-Br-5
`BT-Br-6
`BT-Br-7
`BT-Br-8
`Melanoma
`BT-Me-1
`BT-Me-2
`BT-Me-3
`BA-Me-4
`BA-Me-5
`BT-Me-7
`BT-Me-8
`BA-Me-9
`BA-Me-10
`BA-Me-11
`Colon
`BT-Co-1
`BT-Co-2
`BT-Co-3
`BT-Co-4
`BT-Co-5
`BT-Co-6
`BT-Co-7
`BT-Co-8
`BT-Co-9
`Positive
`control
`Anti-EMA
`MSigG
`Negative
`control
`
`1+
`
`1+
`2+
`
`2+
`2+
`2+
`2+
`
`2+
`1+
`
`1+
`2+
`4+
`
`1+
`1+
`1+
`
`3+
`4+
`4+
`4+
`
`4+
`4+
`1+
`2+
`3+
`
`1+
`3+
`
`2+
`2+
`1+
`
`2+
`3+
`4+
`3+
`1+
`3+
`3+
`1+
`1+
`2+
`
`1+
`
`1+
`
`3+
`
`4+
`
`4+
`
`Positive control
`Negative control
`BT-Br-4
`
`BA-Br-3
`BA-Br-1
`BA-Me-4
`
`BA-Br-5
`
`BT-Ne-3
`
`BT-Co-4
`BT-Co-5
`BT-Co-2
`BT-Co-1
`BT-Co-3
`
`4+
`-
`1+,3+ mostly,
`3 + some areas-
`3 + some areas
`4 + some areas
`4+ some areas
`
`4+
`
`4+, 3+ areas,
`1+,3+CT
`same
`same
`3+, 4+ area, 1 +,
`CT
`
`3 + blood vessels
`
`- mostly, 4+ few
`cells
`4+ blood vessels
`1 +, 2 + some areas
`same
`3+ few areas
`2+
`4+ areas
`- mostly, 2 + few spots 2+, 1 +areas
`- mostly, 2 + few spots 2+, 1 +areas
`1 + areas
`- mostly, 3+, 4+ few
`spots
`
`Tissue sections were tested 1 year apart.
`
`was negative in all-other specimens. Other antibodies,
`such as Me-7 and Me-8 generated from immunization
`with the first tumor biopsy, were present in all speci(cid:173)
`mens. BR-6 was not present in the original specimen
`but was in all subsequent samples, although not in the
`tissue culture line derived from the original lymph
`node. Thus, of the five antibodies specifically derived
`from melanoma tumors, only Me-4, Me-7, and Me-8
`reacted consistently with all the tissues and Me-3,
`and Me-5 had variable patterns. This stresses the im(cid:173)
`portance of having a panel of antibodies and illustrates
`microheterogeneity in multiple biopsies from a single
`patient.
`Table 5 illustrates the results of phenotyping the
`tumor of a patient who had a melanoma removed from
`the right lung (12/86) and, from the left lung (7 /87) who
`also had tissue culture cells grown from an earlier
`lymph node biopsy (2/85). The melanoma lymph node
`typed with Co-l or Co-4 while both lung samples were
`
`Table 7. Antibody typing patterns in breast carcinomas
`
`Patient No.
`
`Br-1 Br-2 Br-3 Br-4 Co-1 Co-2 Co-4 Me-4
`
`19
`20
`24
`15
`16
`1
`5
`10
`22
`12
`Tumor typed
`
`X
`X
`X
`X
`X
`X
`X
`X
`X
`X
`10
`
`X
`
`X
`
`X
`X
`X
`X
`
`X
`X
`X
`7
`
`X
`
`2
`
`X
`
`X
`
`X
`
`X
`
`X
`X
`5
`
`X
`
`X
`
`Tissue culture cells derived from one ofthe lymph node metastases were
`also evaluated.
`Abbreviations: Positive control = anti-HLA class 1 antigens; Anti-EMA =
`anti-epithelial membrane antigen; MSigG = normal mouse immuno(cid:173)
`globulin; Negative control = anti-MSigG, PBS, anti-EMA.
`
`gates were analyzed. All antibodies were against
`membrane determinants except for one patient with
`chronic lymphocytic leukemia who was treated with a
`cocktail of anti-idiotypic immunoconjugates. No sig(cid:173)
`nificant changes were seen in total serum comple(cid:173)
`ment, immunoglobulins, or lymphocyte subsets (data
`not shown). Five patients received less than 500 mg of
`antibody, 4 patients had between 500 and 1 000 mg,
`and 14 patients greatc< than 1.5 g of antibody.
`
`Heterogeneity and selection
`Table 4 shows data from a patient with melanoma.
`Me-3 was positive in the initial lymph node from the
`occipital area and stayed positive in the subsequent
`tissue culture derived cells from that lymph node, but
`
`108
`
`Mol. Biother., 1988, val. 1, no. 2
`
`IMMUNOGEN 2304, pg. 8
`Phigenix v. Immunogen
`IPR2014-00676
`
`

`

`Papers
`
`Table 8. Selection of antibodies for "cocktail"
`
`MoAb
`
`Reactivity
`
`3+, 4+, all over
`
`-, 3+, 4+, CT (2+)
`
`4+ areas, CT (-)
`
`2 +, 3 + areas, CT (-)
`
`Positive control
`Negative control
`BT-Br-4
`BA-Br-3
`BA-Br-1
`BA-Me-4
`BA-Br-5
`BT-Ne-3
`BT-Co-4
`BT-Co-5
`BT-Co-1
`BT-Co-2
`
`Cocktail selected included BA-Br-1 and Br-5.
`Abbreviations: CT = corrective tissue; Br = antibody raised after immu(cid:173)
`nization with breast cancer; Me = antibody raised after immunization
`with melanoma cancer; Ne = antibody raised after immunization with
`renal cancer; Co = antibody raised after immunization with colon
`cancer.
`
`negative. The left lung melanoma metastasis typed
`strongly with Me-11, but the right lung tumor was
`negative. A weak reaction with Br-6 and Br-7 was
`present in tumor in the left lung but not in the right
`lung. This patient confirms the variability of antigenic
`expression on tumor cells within an individual over a
`period of time. At least two variables are involved:
`time and site of metastasis. Among the 10 antibodies
`generated with melanoma immunization all three sam(cid:173)
`ples were positive with six of the antibodies. Lymph
`node metastasis from one patient taken 1 year apart
`were typed and demonstrated relative antigenic stabil(cid:173)
`ity (Table 6). These patients in Tables 4 to 6 illustrate
`that tumor typing may vary with site and/or time and a
`panel of antibodies is needed to approach 100% cell
`coverage.
`Table 7 represents 10 patients with breast carcino(cid:173)
`ma and their typing pattern with eight monoclonal an(cid:173)
`tibodies. The variability of tissue typing of breast car(cid:173)
`cinomas is apparent with monoclonal antibody Br-1
`typing positive for 9 of 10 of the breast carcinomas but
`never with 100% of the cells in a tumor being positive.
`Five of the antibodies typed only one of the breast
`
`Table 9. Flow cytometric analysis of monoclonal anti(cid:173)
`body binding to carcinoma cells
`
`Antibody
`
`Tumor type
`
`%cells
`positive
`
`Peak
`mean
`channel
`
`Type of
`interaction
`
`Ovarian
`carcinoma
`
`Carcino-
`sarcoma
`
`Br-1
`Br-3
`Co-6
`Br-1 + Br-3
`+ Co-6
`Br-1
`Br-3
`Br-5
`Br-1 + Br-3
`+ Br-5
`
`80
`74
`80
`100
`
`82
`37
`12
`59
`
`97
`104
`95
`126
`
`121
`87
`53
`101
`
`Additive
`
`Subtractive
`
`tumors. To ensure typing and saturation a panel of
`monoclonal antibodies is needed.
`Evaluation of results in Tables 4 to 7 demonstrates
`both variability of tumor typing but also certain simi(cid:173)
`larities. Thus, although multiple specimens for tissue
`typing are ideal, a single specimen is most often avail(cid:173)
`able and can give an approximation of tumor typing.
`However, in no situation did a single antibody type all
`cells.
`Table 8 gives an example of the kind of tumor typ(cid:173)
`ing that is done to select antibodies. Br-1 and Br-5
`had the strongest intensity and widest distribution of
`staining, were negative for connective tissue, and
`were picked as the treatment cocktail.
`The histologic evaluation for selection of antibodies
`was complemented by flow cytometry of viable tumor
`cell suspensions reacted with the monoclonal antibod(cid:173)
`ies. Table 9 shows two examples demonstrating both
`an additive effect or negative effect of mixing other
`antibodies. Perhaps hindrance related to the closeness
`of the epitopes or modulation of the epitopes contrib(cid:173)
`uted to the negative interaction.