`
`Z'SH MEDICAL ASSOCIATION TAVISTOCK SQUARE LONDONMWCIH b
`
`,
`
`PFIZER'EX. 1529:;
`
`PFIZER EX. 1529
`Page 1
`
`
`
`Editorial Board
`G Slavin
`(Editor)
`DA McSwiggan (Assistant editor)
`AJ Bellingham
`RAB Drury
`GL Gibson
`NR Grist
`LA Hanson (Sweden)
`Eadie Heyderman
`MRS Hutt
`OM Jones
`HEM Kay
`B Naylor (USA)
`MG Rinsler
`
`AA Sharp
`C Sultan (France)
`IT Whicher
`JS Lilleyman (Publications Secretary,
`Association of Clinical Pathologists)
`Joan F Zilva (representing the
`Association of Clinical Biochemists)
`AI Spriggs (representing the British
`Society for Clinical Cytology)
`RA Thompson (representing the
`British Society for Immunology)
`EDITOR British Medical Journal
`
`Paul Burgon (Technical editor)
`
`General advisory board
`EN Allott
`Rosemary Biggs
`Sir Frank M Burnet (Australia)
`Sir John Dacie, FRS
`WM Davidson
`NH Martin
`Dorothy S Russell
`E Joan Stokes
`MM Wintrobe (USA)
`
`Officers of the Association of Clinical Pathologists
`President: PDJ Holland
`Chairman of Council: AC Hunt
`Vice-presidents: GW Storey
`Honorary Secretary: GW Pennington
`Honorary Treasurer: J Burston
`L Parker
`President-elect: NFC Gowing
`Meetings Secretary: Colin Berry
`Fonner editors: AG Signy 1944-72 HEM Kay 1972-80
`
`Papers should be sent in duplicate to the Editor, Journal of
`Clinical Pathology, BMA House, Tavistock Square, Lon(cid:173)
`don WC1H 9JR. They should be in double-spaced type(cid:173)
`writing on one side of the paper only. The name of the
`author with initials (or distinguishing Christian name)
`should be followed by the name and address of the
`laboratory where the work was performed. A definition of
`the position held by each author should be stated in a
`covering letter to the Editor. All papers with joint author(cid:173)
`ship should be accompanied by the signatures of all the
`authors. Results should not be shown as both tables and
`graphs.
`ACKNOWLEDGMENT A stamped addressed postcard should
`be enclosed for return to the author as acknowledgment of
`receipt of the manuscript. Overseas authors should enclose
`international reply paid coupons.
`ILLUSTRATIONS Diagrams should be drawn in Indian ink
`on stout white paper. The legends for illustrations should
`be typed on a separate sheet. Photographs and photomic(cid:173)
`rographs should be on glossy paper, unmounted.
`CHEMI CAL FORMULAE should as far as possible be written in
`a single horizontal line.
`In general, symbols and abbreviations
`ABBREVIATIONS
`should be those used by the Biochemical Journal.
`Sl UNITS The units in which the work was done should be
`cited first followed by traditional units or SI units in paren(cid:173)
`theses. A conversion factor shall be given as a footnote to
`tables and in legends to figures. Such conversion is the
`responsibility of the author.
`ETHICS The critical assessment of papers submitted will
`include ethical considerations.
`the
`REFERENCES Number references consecutively in
`order in which they are first mentioned in the text. Identify
`references in the text by arabic numerals (in parentheses). ·
`References cited only in tables or in legends to figures
`should be numbered m accordance with a sequence estab(cid:173)
`lished by the first identification in the text of the particular
`table or illustration. The references should be given in the
`form used by Index Medicus:
`Journal (List all authors when six or less; when seven or
`more, list only first three and add et a/.)
`Soter NA, Wassermann SI, Austen KF. Cold urticaria:
`
`release into the circulation of histamine and eosinophil
`chemotactic factor of anaphylaxis during cold challenge.
`N Eng/ J Med 1976;294:687-90.
`Book
`Osler AG. Complement: mechanisms and functions.
`Englewood Cliffs: Prentice Hall, 1976.
`It is the responsibility of the author to ensure that refer(cid:173)
`ences are correctly quoted.
`REPRINTS Twenty-five reprints will be supplied free of
`charge and further reprints may be ordered when the proof
`is returned.
`NOTICE TO ADVERTISERS Applications for advertisement
`space and for rates should be addressed to the Advertise(cid:173)
`ment Manager, Journal of Clinical Pathology, BMA
`House, Tavistock Square, London WC1H 9JR.
`NOTICE TO SUBSCRIBE RS The Journal of Clinical Pathol(cid:173)
`ogy is published monthly. From 1 January 1982 the annual
`subscnption rates will be £45 in the UK and the Republic
`of Ireland and US$110 in all countries overseas. PaymeDt
`for overseas subscriptions should be made in US dollars, ic,
`US$110, or in your national currency based on the prevail(cid:173)
`ing exchange rate to the US dollar of that currency, pay·
`able to the British Medical Association. Orders can also be
`placed locally with any leading subscription agent or boot(cid:173)
`seller. (For the convenience of readers in the USA sub(cid:173)
`scription orders, with or without payment, can also be sent
`to British Medical Journal, 1172 Commonwealth Avenue,
`Boston, Mass 02134 . All enquiries, however, must be
`addressed to the publisher in London.) All
`·
`regarding air mail rates and single copies already
`should be addressed to the publisher in London.
`COPYRIGHT© 1982 Journal of Clinical Pathology.
`publication is copyright under the Berne Convention
`the
`International Copyright Convention. All
`reserved . Apart from any relaxations pennitted
`national copyright laws, no part of this publication may
`reproduced, stored in a retrieval system or transmitted
`any form or by any means without the prior permission
`the copyright owners. Permission is not, however,
`to copy abstracts of papers or of articles on condition
`full reference to the source is shown. Multiple copying
`the contents of the publication without pernussion
`always illegaL
`·
`
`PFIZER EX. 1529
`Page 2
`
`
`
`J Clin Patho/1982;35:369-375
`
`Review article
`Monoclonal antibodies in oncology
`
`KAROL SIKORA
`
`From the Ludwig institute for Cancer Research, The Medical School, Hills Road, Cambridge CB2 2QH
`
`SUMMARY Molecular biology has made tremendous strides over the last five years. The new biology
`allows us to prepare monoclonal antibodies to defined antigens; to detect, isolate and clone
`specific genes; and to insert these genes into defined sites in different cells giving new functions
`to old organisms. These revolutionary developments have been followed closely by research(cid:173)
`ers, businessmen, politicians and philosophers, as well as by those involved in the clinical
`care of patients. Although our understanding of human molecular biology is increasing rapidly,
`it is the development of monoclonal antibodies that has the most immediate appl ication in the
`clinic. There have been several reports of their use in the diagnosis, localisatio n and treatment of
`human malignant disease. This review describes developments that are likely to have direct relevance
`to patient care in the near future .
`
`What is a monoclonal antibody?
`
`The immunological response to any foreign antigen
`is polyclonal: many different clones of B lympho(cid:173)
`cytes are stimulated to produce antibodies. These
`antibodies have different molecular structures and
`in turn recognise different molecular conformation
`patterns on the stimulating antigen-
`the antigenic
`this complexity of antibody
`determinants. It is
`response that makes the antigen-antibody interaction
`difficult to analyse at a molecular level. This is
`particularly so with complex antigens such as the
`tumour cell surface. Monoclonal antibodies occur
`naturally in patients with myeloma. Here neoplastic
`transformation occurs in a clone of B lymphocytes
`with the result that large quantities of identical
`immunoglobulin molecules are produced. It was by
`using myelomas that the chemical structure of the
`immunoglobulin molecule was discovered.! However,
`the antigens to which most myeloma immuno(cid:173)
`globulins are directed are usually unknown and
`are unlikely to be important. In J 975 Kohler and
`Milstein2 constructed a hybrid myeloma (hybridoma)
`which produced a monoclonal antibody directed
`against a specified antigen. Mice were immunised
`with the antigen (sheep red cells) and their spleen
`lymphocytes collected. The lymphocytes were fused
`with an established myeloma
`line and hybrids
`selected by growth in selective tissue culture medium.
`
`Accepted for publication 18 November 1981
`
`The resultant hybrids were rapidly growing (a
`property conferred by
`the myeloma) and yet
`contained new
`immunoglobulin genes (from the
`lymphocytes of the immunised mouse). The hybri(cid:173)
`domas were cloned by diluting the cells and growing
`up colonies from single cells. These cloned hybri(cid:173)
`domas now contained only one set of new immuno(cid:173)
`globulin genes (Fig. J). After growing in
`tissue
`culture the supernatant containing the secreted
`antibody was tested for activity against the im(cid:173)
`munising antigen. Using this system, antibodies can
`be isolated which define single antigens in a complex
`mixture such as the molecules on tumour cell
`surfaces. These molecules can now be compared to
`those appearing on non-malignant cells from the
`same tissue of origin.
`
`Do tumour antigens exist?
`
`There is considerable evidence that the immune
`system responds to antigens on tumour cells, both in
`experimental animal systems and in human neo(cid:173)
`plasia. These tumour antigens are defined by assays
`which utilise the various modes of immune response
`to them. [t is important to distinguish the antigens
`present on the tumour cell surface that are unique
`to tumours and are not shared with normal cells.
`There are several documented examples of such
`antigens within experimental tumour systems.3 4
`Until the development of the monoclona l anti(cid:173)
`body technology, it has been impossible to sort out
`369
`
`PFIZER EX. 1529
`Page 3
`
`
`
`370
`
`Tumour CE'Il surfacE'
`(comp!E'x antigE'n)
`
`Sp!E'E'n CE'!!s
`
`MyE>Ioma
`
`Hybrid
`
`h '(' antibodiE's
`~~
`~
`
`~
`• - o •w
`Fig. 1 Making a monoclonal antibody. A complex
`antigen, such as a tumour cell surface, is used to
`immunise mice. The spleen cells (S) are removed and
`fused with a myeloma line (M) . Hybrids are cloned and
`those antibodies binding to the antigen selected.
`
`the antigenic complexity of human tumour cell
`surfaces. The evidence for the existence of unique
`tumour specific antigens in man analogous to the
`tumour-specific transplantation antigens in animals
`is circumstantial. The natural history of certain
`tumours, the waxing and waning of tumour masses
`the occurrence of spontaneous regression
`and
`suggests that there may be some host control of
`tumour growth.5 Similarly, the relation between
`histological evidence of tumour
`infiltration by
`immunocompetent cells and prognosis suggests that
`these infiltrating cells have some controlling in(cid:173)
`fluence of tumour growth. 6 Further circumstantial
`evidence comes from the increased incidence of
`malignancy in immunosuppressed patients, although
`here the spectrum of tumour types found is not
`similar to that found in the normal population. 7
`Serological analysis and assays of lymphocyte
`function have shown that the immune system in
`
`Sikora
`
`man can actually recognise the tumour cell surface. a •
`Whether immune mechanisms are able effectively to
`destroy tumour cells in vivo remains in question.
`
`Production of monoclonal antibodies to human
`tumours
`
`FUSION SYSTEM
`Currently there are three systems in which anti(cid:173)
`tumour monoclonal antibodies can be raised; mouse,
`rat, and human. For human tumours, mice and rats
`have the obvious advantages of responding to a wide
`variety of antigens and are thus the choice for an
`exhaustive analysis of tumour cell surface com(cid:173)
`ponents. This wide response may be a disadvantage
`in that xenogeneic imrnunisations often result in
`antibodies directed against histocompatibility anti(cid:173)
`gens and blood group substances.
`It is now possible to fuse human lymphocytes
`tumours, either with
`directly from patients with
`mouse or rat myelomas, so obtaining mixed species
`hybrids which produce human monoclonal anti(cid:173)
`bodies. The frequency of hybridisation and the
`quantity of human immunoglobulin produced by
`interspecies hybrids is considerably less than in
`mouse-mouse or rat-rat fusions. A further problem
`is the preferential loss of human chromosomes in
`rodent-human hybrids which results in frequent
`loss of immunoglobulin production. There are now,
`however, several human myeloma lines available
`which are suitable for fusion.1o u Such lines must
`be rapidly growing and have an appropriate genetic
`selection mechanism to enable the parent myeloma
`to be killed in the hybridoma mixture. Once estab(cid:173)
`lished, human-human hybrids show no apparent
`preferential loss of chromosomes and thus the
`stability of the hybrid is assured. The quantity of
`immunoglobulin secreted by these human-human
`hybrids is usually of the order of 1 P-gfml which is
`one tenth of the output of the corresponding mo~
`hybridoma system. There are several advantages tn
`using human lymphocytes to produce monoclonal
`antibodies. The spectrum of the human imm~e
`response which serologically defines tumour-specific
`antigens can be examined. There is abunda~t
`evidence that patients with cancer at some time 1.n
`the natural history of the tumour have in thelf
`serum antibodies which recognise their own tu(cid:173)
`mours.12 The titre of these antibodies is low an~~
`far there have been no good studies on the chenu .
`nature of the determinants recognised by such antlj
`bodies. By obtaining the antibodies in monoclo;a
`form and in sufficient quantity such chemical st~ 1~
`are possible. Lymphocytes from cancer patie~
`can be collected from several sites. Peripheral bl~ g
`lymphocytes may not represent a good startJD
`
`PFIZER EX. 1529
`Page 4
`
`
`
`IIIAJ'"'"'v" from which to perform fusions. More
`to be involved in antitumour activity are the
`illnr~hc1cvtes in the lymph nodes draining a tumour.
`can easily be collected in large
`kJllti1:ies from patients with breast, lung and colo-
`cancer. Another source of lymphocytes for
`comes from
`the tumour itself. Certain
`for example gliomas, are often heavily
`llltrat<:O by lymphocytes. These lymphocytes can
`collected, separated from the tumour and fused
`a human myeloma line.l3
`
`NISATION SCHEDULE
`xenogeneic immunisations the choice of anti(cid:173)
`material and the schedule in which it is used
`immunisation has varied considerably. Very
`detailed work has been performed on opti(cid:173)
`these schedules. Sources of tumour material
`immunisation can come from cell lines grown
`vitro, pieces of fresh tumour tissue, membrane
`from fresh
`tissue, or fractionated
`components from fresh
`tumour cell
`These different
`immunisation pro(cid:173)
`will almost certainly result in different
`of antibodies.
`the production of human monoclonal anti(cid:173)
`immunisation is not possible and the choice
`in the source of lymphocytes for fusion. There
`as yet no evidence to suggest that any particular
`of lymphocytes- peripheral blood, spleen,
`node or intratumour- results in a higher
`!tq1uer1cy of the required antibodies.
`
`--.~~ ' ""G METHODS
`production of antibodies against human tumour
`surfaces requires the screening of many fusion
`to find suitable immunoglobulins. Several
`Rateg;ies have been developed. The commonest
`is to immunise mice with a chosen tumour
`line, for example a melanoma. The fusion
`are screened on that melanoma in an
`binding radioimmunoassay (see Fig. 2) and
`activity of any positive supernatants determined
`other melanomas as well as on cell lines of
`types, both normal and malignant (Fig. 3).
`way the specificity of the monoclonal anti(cid:173)
`is characterised and its ability to distinguish
`cells from their normal counterparts is
`at'!rm,in •• ,.l
`Screenin~ can also be performed using primary
`material. Membrane preparations of tumours
`be used to immunise rodents; the same mem(cid:173)
`preparation can be bound to plastic wells and
`in a solid phase radioimmunoassay to screen
`activity of resulting monoclonal antibodies.
`variant of this screening procedure is to use
`
`371
`
`Rad iolabell~
`anti lg
`
`Monoclonal
`antibody
`
`l
`
`f Cl 0 Cl
`
`ODD
`Tumour cell membrane
`
`Direct
`I ndirect
`Fig. 2 Binding assays for monoclonal antibodies. In
`the indirect assay bound monoclonal antibody is detected
`by a radio/abe/led anti-immunoglobulin. In the direct
`assay internally labelled-for example, 3H-lysine,
`antibody is used.
`
`Cell line
`Fresh tumour
`Tumour membrane
`Solubil ised fractions
`
`Human lymphocy tes
`Peripheral blood
`Spleen
`Lymph nodes
`lntmtumoural
`
`Spleen cells
`
`Fusion with ___
`
`.....J
`
`1 Mouse
`l
`L
`myeloma l
`/T~
`
`Radioimmunoasm
`
`lmmunohsto!Qgy_
`
`J;ytotoxicity_
`
`Complement
`lmrrunofluorescence
`Membranes
`lmrrunoperoxidase
`Whole cells
`K cells
`Fig. 3 Strategies for making and screening monoclonal
`antibodies to human tumour antigens.
`
`sections of normal and tumour material to look at
`the activity of monoclonal antibodies histologica11y
`by immunofluorescence on frozen sections or by an
`immuno-peroxidase technique. This latter technique
`has the advantage of allowing retrospective surveys
`of paraffin block material readily available from
`hospital pathology departments. By comparing
`tumour samples from different patients with cancer
`
`PFIZER EX. 1529
`Page 5
`
`
`
`372
`
`of the same or different tissues, additional infor(cid:173)
`mation of diagnostic value can be sought.
`In the production of human monoclonal antibodies
`screening strategies are less well worked out. One
`problem is
`the ubiquitous presence of variable
`amounts of human
`immunoglobulin
`in human
`tumours. The detecting anti-human Ig, whether
`fluorescein coupled or radiolabelled, binds to this
`resulting
`in high background
`levels
`in
`tumour
`membrane preparations. This can obscure binding
`by relatively small amounts of high affinity mono(cid:173)
`clonal antibody. This problem can be overcome by
`using cell lines for screening although of course this
`results in selection. A more laborious technique is
`to
`radiolabel
`internally each human
`immuno(cid:173)
`globulin produced by the hybrids by incorporating
`a radioactive amino acid such as 3H-Iysine and screen
`in a direct binding assay.14
`
`Antitumour monoclonal antibodies currently
`available
`
`COLO RECT AL CA RCI NOMA
`Colorectal cancer is a common problem in clinical
`oncology. Diagnosis
`is often difficult, requiring
`extensive endoscopic or radiological investigation.
`The assessment of recurrent disease
`following
`primary surgery is usually impossible until large
`masses of neoplastic tissue have accumulated. For
`the last 15 years much effort has been spent investi(cid:173)
`gating carcinoembryonic antigen (CEA), an antigen
`detected by an antiserum produced in rabbits after
`immunisation with extracts from colonic cancer.
`This antigen , a glycoprotein with a molecular weight
`of 180 000,
`is found
`in several gastrointestinal
`tumours, some lung and breast tumours as well as in
`normal fetal colon.1 5 Considerable interest has been
`aroused in the possibility that the measurement of
`CEA in the blood would relate to the tumour load in
`an
`individual patient,
`thus producing both a
`diagnostic test and a marker for monitoring progress
`of the disease. A major problem in the use of CEA
`for these purposes has been the extensive cross(cid:173)
`reaction between CEA and a variety of similar
`glycoproteins such an non-specific cross-reacting
`antigens (NCA), biliary glycoprotein (BGP) and a
`glycoprotein found in washings of normal colon
`(NCW) . These glycoproteins share antigenic deter(cid:173)
`minants with CEA and
`therefore confuse
`the
`serological analysis since different
`immunisation
`and absorption protocols result in the production
`of different antibodies in the resulting antiserum.
`Monoclonal antibodies give more precise informa(cid:173)
`tion about the interrelations between these cell
`surface components and thus lead to more selective
`and sensitive assays
`for
`truly
`tumour-related
`
`Sikora
`
`products.
`Several groups have now produced monoclonal
`antibodies to either CEA or other antigens present
`on colorectal carcinomas. So far these attempts have
`been carried out by · immunising mice with either
`purified CEA preparations or colorecta l carcinoma
`cell lines. After fusion, screening on either CEA or
`the immunising cell line has been used to identify
`interesting supernatants. Accolla and his colleagues!&
`raised 400 hybrids from mice immunised with puri(cid:173)
`fied CEA and found two which secreted antibodies
`reacting specifically with
`two different a ntigenic
`determinants present on CEA molecules. The
`affinities of these antibodies are relatively high and
`could be used
`to characterise solubilised CEA
`immunochemically. Herlyn et a/17 immunised mice
`with cells grown in vitro. The screening assays
`included radioimmunoassay, mixed haemabsorpt ion
`assays and immunofluorescence on the immunising
`cell line. Two hybridomas were found which secreted
`antibodies binding specifically to human colorectal
`carcinomas, either growing in culture or obtained
`directly from patients. These antibodies do not bind
`to other malignant
`to normal colonic mucosa,
`cells or to CEA.
`
`MELANOMA
`Melanoma is a tumour studied frequently by im·
`munologists. Serology, using panels of patient sera
`and melanoma cells, has been used to construct
`large serological matrices. The biochemical separ(cid:173)
`ation of the different serologically recognised antigens
`has been hampered by the low titres of the sera.
`There are several monoclonal antibodies against a
`variety of human melanoma antigens. Some of these
`antibodies are directed against the human DR
`locus-related) antigen . In one study,18
`(HLA D
`three out of six hybridoma secreted antibodies were
`found to bind to the majority of melanoma cell
`lines and to astrocytomas, as well as to all normal
`and Epstein-Barr virus-transformed
`lymphocytes
`tested (the same distribution as the D R antigen).
`Two of the remaining antibodies, however, were
`found to detect two different antigens common to
`melanoma and astrocytoma cells only. T he m~st
`elegant analysis of the use of monoclonal antibodtes
`in characterising antigenic systems on the surface
`of human tumours comes from the work of Dippold
`and his collaborators.19 Mice were immunised with
`the melanoma cell line SK-MEL 28 and the 1 ~
`antibodies derived were tested on a large panel 0
`human cell lines from a variety of tumour types, as
`well as on early cultures of normal tissue. Se_r~
`logical studies, in conjunction with immunoprectP~
`tation analysis of radiolabelled cell extracts an
`antibody inhibition tests with solubilised a ntigens
`
`PFIZER EX. 1529
`Page 6
`
`
`
`- .11'-<l''"'" that the 18 monoclonal antibodies recog(cid:173)
`six antigenic systems. Two of the systems are
`with molecular sizes of 95 000 and
`
`Commercially available monoclonal antibodies to
`lymphocyte differentiation antigens
`
`Antibody
`
`Reactive populations
`
`373
`
`T CANCER
`(eri!Oil:entel·c monoclonal antibodies have been raised
`breast tumour lines, although the number of
`available is less than in colorectal and
`systems. A monoclonal antibody that
`have considerable clinical use is that raised
`the human oestrogen receptor. 20 It is known
`the presence of oestrogen receptors in breast
`tissue is an indicator of the likelihood of
`hormone treatment. The derivation of
`llOttloc:IOJJal antibodies which can be used for im(cid:173)
`detection of receptors would
`the pathologist's ability to provide
`w"•u••a .. on of prognostic value to the clinician .
`using
`lymphocytes derived
`from axillary
`nodes from patients with breast cancer,
`immunoglobulins which bind
`to breast
`noma cells have been produced.21 A human
`monoclonal antibody produced in this way has
`shown
`to discriminate between mammary
`cells and normal mammary epithelial
`This antibody also reacted significantly with
`,_ •. K.,, ... .,·,.., mammary carcinoma cells
`in
`lymph
`of breast cancer patients with no binding to
`normal lymphocytes or to the stroma of the
`node.
`
`PHOMA AND LEUKAEMIA :
`wide range of monoclonal antibodies has been
`against myeloid and lymphoid neoplasms.
`lymphocytes with different biological
`ions-for example, helper and suppressor
`on antibody synthesis, can be distinguished
`·
`surface markers. Not surprisingly neo(cid:173)
`transformation in cells of the lymphoid
`ts in the clonal expansion of a population
`s bearing a distinct surface marker pattern.
`conventional serology such patterns have
`been related to prognosis as in the sub(cid:173)
`of lymphatic leukaemia into T, B and
`ALL types. With monoclonal antibodies a
`finer discrimination can be made and used to
`therapeutic approaches to these diseases. 22
`range of monoclonal antibodies available to
`lymphoid subpopulations is outlined in the
`It should be stressed that these antibodies do
`recognise tumour antigens but clonally expanded
`antigens.
`
`helper/inducer T cells, certain leukaemias,
`mycosis fungoides, S~za ry syndrome
`
`peripheral blood T lymphocytes. T cell
`leukaemia, mycosis fuogo ides
`
`*OKT 3
`t LI7 Fl2
`t NEI-01 6
`*OKT 4
`t SK 3
`t SK 4
`thymic lymphocytes, some thymomas
`*OKT 6
`suppressor/cytotoxic T lymphocytes,
`*OKT 8
`certain T cell neo pl asms
`t SK I
`immature T cells, certain leukaemias
`*OKT 10
`anti HLA, DR (Ia) B lymph ocytes a nd B cell neo plas ms
`anti lg
`(nodular lympho ma, most chro nic
`lymphocytic leukaemias, myelo ma)
`
`Commercial suppliers:
`*Ortho Ph armaceuticals, Denma rk St, High Wycombe, Bucks HPII
`2ER .
`t Becto n Dickinson, 490-3, La keside Drive, California 94086 USA.
`t New England Nuclea r, 2 New R oad, Southampton S02 OAA .
`
`OTHER TUMOURS
`Monoclonal antibodies have been or are being
`raised against a wide variety of human tumours,
`including gliomas, neuroblastomas, sarcomas, lung
`cancer as well as bladder, prostrate and testicular
`tumours.
`
`Clinical uses
`
`DIAGNOSIS AND MONITORING
`A major problem in clinical oncology is the measure(cid:173)
`ment of tumour load in an individual patient.
`Less than 10 % of all cancer patients have disease
`which can be reliably assessed by conventional
`techniques, such as palpation or diagnostic radiology.
`This hampers the evaluation of different treatment
`methods. Certain relatively rare
`tumours shed
`products into the circulation; and the concentration
`of these tumour markers can be related to the total
`tumour cell burden. Examples include a-fetoprotein
`in hepatoma and
`teratoma; human chorionic
`gonadotropin
`in choriocarcinoma and CEA in
`some colorectal carcinomas. Other tumour-related
`molecules are also shed into the serum but until
`now there has been no way of detecting them. By
`using specific monoclonal antibodies in a suitable
`radioimmunoassay, picogram quantities of these
`shed products can be measured. A large panel of
`well characterised monoclonal antibodies will
`therefore have considerable diagnostic use at several
`the management of cancer patients.
`stages
`in
`Firstly, patients presenting with symptoms sug(cid:173)
`gestive of malignancy may have no tissue readily
`accessible for biopsy. Investigations now necessary,
`are often expensive, time-consuming and cause the
`patient considerable discomfort. Early carcinoma of
`the pancreas is a good example. The second use of
`
`PFIZER EX. 1529
`Page 7
`
`
`
`374
`
`monoclonal antibodies is for regular screening in
`patients with conditions that are known to pre(cid:173)
`dispose to neoplastic changes, such as ulcerative
`colitis, polyposis coli and certain forms of hepatic
`cirrhosis. Thirdly, the serial monitoring of tumour
`marker concentration in an
`individual patients'
`serum could provide a reliable index of the behaviour
`of the tumour and its reponse to treatment. There are
`several reports of monoclonal antibodies being used
`to detect circulating tumour markers. A monoclonal
`antibody detecting a monosialoganglioside from
`colo rectal tumour cells has been used to screen serum
`samples.23 Blocking activity was found in the serum
`of 24 out of 32 patients with colorectal cancer but was
`not present in the serum of 38 healthy donors and 36
`patients with other cancer types. This sort of observa(cid:173)
`tion has now been made with several different anti(cid:173)
`bodies to a variety of tumour types.
`
`HISTOLOGICAL EVALUATION
`Recent advances in oncology have made the therapist
`more dependent on his pathologist colleagues than
`ever before. The oncologist is only too familar
`with
`the problems occuring
`in distinguishing
`certain anaplastic carcinomas from
`lymphomas
`and even teratomas. Immunohistology with mono(cid:173)
`clonal antibodies can provide considerable diagnostic
`information. In a study of 33 cases of non-Hodgkin's
`lymphoma by a panel of monoclonal antibodies to
`different lymphocyte subpopulations information
`was provided that was unobtainable by conventional
`microscopy.24 Using such techniques more can be
`learnt about the different types of lymphoma and
`their response to treatment. Antibodies are also now
`available which can discriminate between different
`histological types of common solid tumour; an
`example is an antibronchial carcinoma antibody
`which binds only to smaJJ cell tumours.25 Such
`reagents will be of great value when limited amounts
`of material are available for pathological examination
`as with sputum cytology. A series of monoclonal
`antibodies to various tumour types would be of
`considerable use in evaluating histological material
`from patients presenting with metastatic disease.
`In this way treatable forms of cancer could be
`excluded without the costly and uncomfortable
`exercise of hunting the primary tumour.
`
`TU MOUR LOCALISATION
`The use of radiolabelled polyclonal antitumour
`antibodies for tumour detection has been attempted
`with limited success. The major problems have been
`the lack of a suitable antibody giving sufficient
`target to non-target contrast for imaging and also
`in reproducibility preparing and
`the difficulty
`purifying antitumour antibodies. The development
`
`Sikora
`
`of monoclonal antibodies has several advantages.
`their defined specificity may allow the
`Firstly,
`contrast required for effective imaging and, secondly,
`their production is reproducible on a large scale.
`Furthermore, they represent a concentrated form
`of immunoglobulin with defined activity and thus
`the total foreign protein load given to an individual
`patient is much lower. It has been shown that mouse
`monoclonal antibodies to CEA and to teratocarci(cid:173)
`nomas can localise human tumour deposits
`in
`immunosuppressed mice bearing human tumour
`xenografts.26 27 There are as yet, however, few
`reports on the use of labelled monoclonal antibodies
`for localisation in man. Mach and his colleagueszs
`injected 14 patients with large bowel and pancreatic
`cancer with 131J-labelled purified mouse monoclonal
`anti-CEA antibody. In eight patients
`increased
`radioactivity was observed in the region of the known
`tumour deposit. To detect blood pool and secreted
`radioactivity a 99mTechnetium scan was also per(cid:173)
`formed. After subtraction of 99mTc radioactivity
`from 131I radioactivity a two to tenfold concentra(cid:173)
`tion of 131I activity was found in the tumour sites.
`
`TREATMENT
`Monoclonal antibodies are currently being used by
`several groups in attempts to assess their value as
`therapeutic agents. A T ceJJ specific murine hybri(cid:173)
`doma monoclonal antibody has been infused into
`patients with T cell neoplasms.29 The antibody used
`reacts with normal human T cell differentiation
`antigens which are present in increased amounts on
`the malignant cells of patients with cutaneous T cell
`lymphomas. Favourable but temporary responses
`were seen in patients with T cell leukaemias and
`with mycosis fungoides, a neoplasm of T helper
`lymphocytes. In another studyao a mouse monoclonal
`antibody directed against a lymphoma-associated
`antigen was given to a patient with lymphosarcoma
`cell leukaemia. Transient decreases in circulating
`tumour cells and the appearance of dead cells were
`noted after the infusion of 75 mg of antibody. The
`radiation
`patient had received extensive prior
`therapy and chemotherapy. There are several mecha(cid:173)
`nisms by which this tumour cell destruction can
`occur. These include the activation of complement;
`the triggering of antibody-dependent cell-mediated
`cytotoxicity; and opsonisation resulting in macro(cid:173)
`phage killing. It is probable that such mechanisms
`alone are unlikely to destroy large tumour masses
`but could d