PRE-TARGETED RADIOIMMUNOTHERAPY WITH

STREPTAVIDIN-CC49 MONOCLONAL ANTIBODY AND 'Y-DOTA-BIOTIN

Rommel J. Domingo

A thesis submitted in confonnity with the requirements

for the degree of Master of Science

Graduate Department of Phmaceutical Sciences

University of Toronto

8 Copyright by Rommel J. Domingo 199'7

National Library 1*1 of Canada Bibliothéque nationale du Canada

Acquisitions and Acquisitions et Bibliographie Services senrices bibliographiques

395 Wellington Street 395. rue Wellington OttawaON K I A ON4 Ottawa ON K1A O N 4 Canada Canada

Your Me Voire mrérmw

Our Ns Nolm trefdrenu,

The author has granted a non- L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or seil reproduire, prêter, distribuer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/film, de

reproduction sur papier ou sur format électronique.

The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fi-om it Ni la thèse ni des extraits substantiels may be printed or othenirise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation.

PRE-TARGETED RADIOIMMUNOTHERAPY WITH STREPTAVIDIN-CC49 MONOCLONAL ANTIBODY AND oY-DOTA-BIOTIN

Master of Science 1997, Rommel J. Domingo Department of Pharmaceutical Sciences, University of Toronto

ABSTRACT

In this study pre-targeted radioimmunotherapy (RIT) of ath ymic mice bearing

subcutaneous LS 174T human colon cancer xenografts was evaluated using streptavidin-CC49

and yttrium-90 (goy)-DOTA-biotin. The biodistribution of DOTA-biotin, labeled with indium

1 1 1 ( 1 b), was first investigated in BALBIc mice to determine i ts suitabili ty for pre-targeting

strategies. Indium 1 1 1-DOTA-biotin was rapidly eliminated frorn b l d and normal tissues. A

dosing schedule was then determined by evaluating different time intemals between the

administration of streptavidin-CC49 and DOTA-biotin and doses of DOTA-biotin in mice

bearîng LS174T colorectal senografts. Cornparison of 40 and 77 hour time intervais indicated

that the 40 hour time interval would be the most effective for the pre-targeted RIT study. A

cornparison of 1, 30, 40, and 150 pg doses of DOTA-biotin indicated that 40 pg nearly

achieved maximal uptake of DOTA-biotin in the turnour. Using the dosing schedule developed,

the therapeutic efficacy and toxicity of pre-targeted RIT wilh 900 pCi of gOY-DOTA-biotin was

evaluated in athymic mice bearing LS 174T colorectal xenografts. Considerable suppression in

turnour growth was observed in 1/6 treated mice. N o apparent bone marrow toxicity and

deterioration in health were observed in treated mice. In summary. pre-targeted RIT with

streptavidin-CC49 and 9V-DOTA-biotin demonstnted some promise as an effective and sale

approach in the treatment of colorectal cancer.

ACKNOWLEDGEMENTS

1 would like to extend my sincere gratitude to Professor Ray Reilly whose guidance,

encouragement, and kindness have made the past two years a very rewarding and mernorable

experience. Moreover, his faimess, approachabili ty , and outgoing nature made worhng at The

Toronto Hospital very cornfortable and enjoyabie. 1 will always be grateful to Professor Reilly.

for his teaching and exemplary research have helped me accomplish my educational goals.

1 would also like to thank the staff, past and present, of the Radiopharmacy Department

Division or Nuclear Medicine at the Toronto Hospitai, particularly those who helped me in

rny research and those who provided the many laughs and fun times. Much appreciation also

goes out to my Advisory Cornmittee; Prof. B. Berry, Prof. B. Bowen, and Prof. J. Ballinger.

Thank-you for your guidance! Lastly, thanks to my parents, Danny and Tonette, Ritchie,

Nadia, Gina. Nick, and Dave. Your deep pocket$, support, and guidance will always be

remem bered.

TABLE OF CONTENTS

Page

LIST OF TABLES

LIST OF FIGURES

LIST OF ABBREVIATIONS

LIST OF APPENDICES

INTRODUCTION

RADIOIMMUNOTHERAPY CONCEPTS Factors Influencing Tumour Uptake of Antibodies

Considerations for Radioimmunotherapy

A~ltibody Propuries

Radiori ztclides

Metabolism of Rndioncrclides

Radio bio bg y

Limitations of Radioimmunotherapy

S trategies to 1 mprove Radioimrn unotherapy

Irnproving Tumoar Llptuke of Anribodies

Enlioncing Tz~miir Cytotoxicity

Rediicirig HAMA

Redrici~ig Bone Mavo w Toxiciiy

Page

PRE-TARGETING WITH THE (STREPT) AVIDIN-BIOTIN SYSTEM Propertîes of (Strept)avidin and Biotin

B iotinylated mAb and Radiolabeled Streptavidin (-+tep)

S t r e p ~ ~ i d i n y lated mA b and Radiolabeled Bioiin (2-step)

Biotinylated mAb, (Strept)avidin, and Radiolabeled Biotin (3-step)

Pre-targeting with Streptavidin-CC49 Monoclonal Antibody and DOTA-biotin

CC49 Monocloml Antibodv

Streptuvidi~tviaiion 01 CC49 Mo~toclortal &iibody

DOTA-biotin

RESEARCH OBJECTIVES Statement of the Problem

Objectives of Research

EXPERIMENT 1: BIODISTRIBUTION OF DOTA-BIOTIN Introduction

Results

Discussion

Page

EXPERIMENT II: DEVELOPMENT OF A DOSING SCHEDULE FOR PRE-TARGETED RADIOIMMUNOTHERAPY WITH STREPTAVIDIN-CC49 MONOCLONAL ANTIBODY AND 9 ('Y-DOTA-BIOTIN In traduction

Results

Discussion

EXPERIMENT III: PRE-TARGETED RADIOIMMUNOTHERAPY WITH STREPTAVIDIN-CC49 MONOCLONAL ANTIBODY AND 9 'Y-DOTA-BIOTIN Introduction

Resul ts

Discussion

CONCLUSIONS AND FUTURE PERSPECTIVES

REFERENCES

LIST OF PUBLICATIONS AND ABSTRACTS

APPENDICES

LIST OF TABLES

Page

Summary of various animal radioirnrnunotherapy trials 4

Summary of various clinical radioimmunotherapy trials 5

Radionuclides sui table for mdioimmunotherapy 17

Tissue distribution of DOTA-biotin 54

Biodistribution of s~eptavidin-CC49 using a 40 or 72 hour time intend between adrnînistration of ( 1 2 9 - s t r e p t a v i d n - 4 and 1 1 1 ln-DOTA- biotin 66-67

Turnoucnormal tissue ratios of sueptavidin-CC49 using a 40 or 72 hour time intepal between admimsuation of (l'jI-~treptavidin)-CC49 and " ' ln-DOTA-biotin 68

Biodistribution of DOTA-biotin follouing pre-targeting with suepta~idin- CC49 usine a 40 or 72 hour time intemal 70-7 1

Tumour normal tissue ratios of DOTA-bioti n follon-ing pre-targeting wi th streptavidin-CC49 using a 40 or 72 hour ti me intend 73

Individual tumour volumes of LS 174T senografts follouing treatment with strepta~idin-CC49 and g%DOTA-biotin 89

LIST OF FIGURES

S tmcture of IgG antibody

Generation of mAb fragments, chimeric mAb, and humanized mAb

Schematic representation of the three variations to the (strept)avidin-biotin pre-targeting sys tem

Structure of 1, 4, 7, 10-tetraazacyclodoàecane-N, N', N", N"'- tetraacetic acid (DOTA-biotin) developed by the NeoRx Corporation

Elimination of DOTA- biotin from blood

Tissue distribution of DOTA-biotin

Size-esclusion high pressure liquid chrornatography analysis of CC49 mAb purified frorn mouse ascites fluid

Sodium dodecyl sulphate-polyacrylamide gel electrophoresis of CC49 mAb and streptavidin-CC49

Size-exclusion high pressure 1 iquid chromatography anal ysis of streptavidin-CC49

Tumour uptake of DOTA-biotin at increasing doses

Turnour growth curves following pre-targeted radioirnmunotherapy

Suppression of LS 174T turnour growth lollowing pre-targeted radioi mmunotherapy

Evduation of 9 OY -DOTA-biotin tosicity following pre-targeted mdioimmunotherapy

Page

8

10

35

45

52

53

63

63

64

73

86

87

90

LIST OF ABBREVIATIONS

ip IL- 1 1 L-3 Ka kDa mAb MRUs P' w RIT SDS-PAGE sui fo-SMCC TAG-72 US FDA WBC XRT

copper 67 yttrium 90 indium 11 1 iodine 125 iodine 131 rhenium 188 astatine 2 1 1 bismuth 2 12 alpha

gamma autologous bone marrow transplantation bovine su bmaxillay rnucin cornplementari ty determining reg on arcinoembryonic antigen colony stimulating factors 1, 4, 7, 1 0-tetraazacyclododecane-N, N'. N", Nu'-tetmacetic acid diethyleneuiaminepentaacetic acid extracorporeal immunoadsorption grays hurnan anti-mouse antibody human anti-streptavidin an6body high pressure liquid chrornatography i ntravenous intrapen t o n d interleukin- 1 interleukin-2 binding affinity constant k i lodd tons monoclonal anti body molecular recognition units p s t 4 nj ec tion reticuloendotheIiai system ndioimm unotherapy sodium dodecyl sulphate-polyacrylamide gel electrophoresis sul fosuccinimidyl-4(NNmaleimideorneth y i ) c y c l o h e e - 1 -carboxylate tumour-associated glycoprotein 71 United States Food and Drug Administration white blood cells extemal radiation therapy

LIST OF APPENDICES

Page

1. Purification of CC49 Monoclonal Antibody From Mouse Ascites FIuid 112

I I . Preparation of S treptavidin-CC49 Monoclonal Antibody 113

II 1. Measurement of the Lmmunoreactive Fraction of Streptavidin-CC49 Monoclona1 Anti body 114

IV. Preparation of 1 1 1In- and goy-DOTA-biotin 115

V. Establishment of LS 174T Colon Cancer Xenografts 116

VI. Caiculation of Absolute Streptavidin-CC49 and DOTA-biotin Upiake in LS 174T Xenografts 117

INTRODUCTION

Radioimmunotherapy (RIT) is a form of cancer treatment in which radiolabeled

antibodies are administered to a patient t o selectively deliver ionizing radiation to the lesion.

The idea is t o deliver a sufficient radiation dose to the tumour tissues in order to achieve

suppression of tumour growth, or at best, endication of the tumour, while normal tissues are

spared tosicity. Radioantibodies are selectively taken up in tumour tissues by binding to

specific antigens o r receptors located on the surface of the cancer cells. These antigens are not

tmly specific to cancer cells, but nther, a re tumour-associated. They are highly overespressed

in tumour tissues compared to normal tissues, which in most cases, provides for a sufficient

gradient that achieves anti body locdization in the tumour. Many anti- turnour antibodies are

pan-carcinomic, capable of recognizing more than one tumour-associated antigen. In many

cases, a given cancer c m also be targeted with more than one type of anti body.

The earliest reports of using antibodies to treat cancer were published in 1895 by

Hericourt and Richet (Hericourt and Richet, 1895) and in 1906 by Paul Ehrlich (Himmel wei t.

1957). These earl y attempts, however. were neither successful or reproduci ble, and the

properties of the antibodies were unknown. The idea, however, of treating cancer with

antibodies was launched. In the late 1940's a method to label proteins with iodinc 13 1 (1311)

was developed (Eisen and Keston, 1950) and one decade later. clinical trials of RIT bcgan.

Beienvai tes (Beienvaltes, 1974) administered l 3 1-labeled nbb i t antibodies to 14 patients ivi th

metastatic melanoma and achieved a pathologically documented response in one patient.

Thereafter, polyclonal antibodies against turnour-associated antigens such as, the prototypical

carcinoembryonic antigen (CEA), were developed, used in patients. and produced some

favounble responses (Ettinger et al., 1979; Ettinger et al.. 1982). The field of RIT was

revolutionized in 1975 when Kohler and Milstein (Koh l e r and Milstein, 1975) introduccd

hybridoma technology, a method that produces large quantities of murine monoclonal

antibodies (mAbs) with high. pre-defined specificitg. Since then, nurnerous mAbs against a

wide variety of human cancers (Reisfield and Cheresh, 1987) have k e n developed and

correspondingly, research efforts in RIT have increased trernendously.

Since 1975 hundreds of papers dealing with RIT have been published. More than

twenty years of intense research in the field have yielded important findings, which a n be

summanzed as follows (Stigbrand et al., 19%): i ) the arnount of radiolabeled antibody

localized in solid tumours of patients is very small at 4 . 1 Q of the injected activi ty : i i ) as a

result of the low antibody uptake in turnours, the total absorbed radiation dose delivered to the

turnour is also very smail. typically C gray (Gy): iii) increasing the administered activity in

order to increase the radiation dose delivered to the tumour is limi ted by bone rnarrow toxici ty,

caused primarily by passive irradiation of the bone marrow by the circulating ndioantibodies;

iv) the immunogenicity of murine mAbs triggers the production of human anti-rnouse

antibodies (HAMA) in patients, thereby limiting the number of treatmcnts theÿ c m undergo; v)

significant tumour regression or complete responses are often achieved in animal models, while

mostly partial and transient responses are achieved in patients with cenain solid turnours; and

vi) hcmatological rnaiignancies such as, B-cell Iyrnphornas, are generally more responsive to

RIT than certain solid turnours.

Monoclonal antibodies themselves are attractive candidates for cancer thenpy because

of the specificity the) can have for tumour tissue. They are considered to be relatively salé,

wi th tosici ty limited on1 y to transien t irnmune-rnediated or allergic- typc reactions (Dillman et

al., 1986). Allergic reactions such as hives are uncornmon and severe anaphylasis is rare

(Kuzel and Duda, 1992). Other reactions include, lever, chills, dyspnea. hypotension, and

shonness of breath (Goldenberp. 1994; Kuzel and Duda, 1992). Currently. there is no mAb

preparation approved for clinical RIT. Two mAb products, however, Satumomab pendetide

(Oncoscint CR/OV@) and Arcitumomab (cEA-sc~~@) have been approved by the United

States Food and Drug Administration (US FDA) for diagnostic radioimmunoirnaging OF

prostate and colorectal cancer, respectively. Radioimmunoimaging is a cancer irnaging modaliy

analogous to that of RIT, in which radiolabeled mAbs are targeted to turnours, that arc

subsequently visualized by a gamma canera. Radioimmunotherapy has suffered in that only

sub-therapeutic doses of radiation are achieved in the tumour, while high whole body

irradiation has regularl y resul ted in bone marrow toxici ty and correspondingl y, poor

therapeutic indices. Moreover, the immunogenicity of murine mAbs precludes repeated courses

of therapy, since HAMA responses can significantl y al ter the biokinetics of the adrninistered

radioanti body. Prornising results have been obtained from animal RIT studies (Table 1).

however, clinical results have yet to demonstrate a similar degree of success (Table 7).

Advancements in rnolecular biology and in radiochemistry, d o n g with the developrnent or

innovative tumour targeting methods are paving the way for new and alternative stntegies that

can potentially overcorne the obstacles currently limiting the clinical success of RIT. For

csample, one innovative strategy to reduce bone rnarrow toxicity involves pre-targeting, in

which a multi-step approach utilizing (strept)avidin and biotin is employed (see: "Prc-targeting

with the (Strept)avidin-biotin System"). Reducing bone marrow toxicity is key io improving

the clinical success of RIT because it may also allow larger arnounts of ndioactivity to be

adrninistercd to the patient, which in tum, may possibly lead to improved radiation doses

delivered to tumour tissues.

Table 1. Summuy of' vÿnous animal radioimmunotherüpy iriüls.

Cancer Cell Line mAb; Radionuclide Administered Therapeutic Toxicity Reference Type Specificity Dose, Route Responses

Mammary Obîained from B W495136 1311 . . . 300 [ici, i v Mean decrease of 4 0 % decrease in Senekowitsch patient biopsy

Mammary SKOV3, 4D5,7C2; 1311 and ovarian SKBR3

400-700 ciCi, i v ;in ti-HER- Yrreir onco- protei n

T-=Il AKWJ, S M , 3 1 E6.4; 1311 1500 pCi, iv lymphoma AKRICum Si. 1 anti-Thy- 1.1

50% and 88% in tumour diameter and tumour volume, respectively Mmked i nhi bi tion of iumour grow th

A bsence of palpable tumour in 7 1 % of mice Up to 7 1 <70 inhi bition of iumour grow t h ai 50 pCi dose

93% reduclion in tumour mass

98% reduction in tumour grow th rütc

leukwyte md et al., 1989 platelet counts

MyeIosuppression De Santes ei al., 1992

1 O()% of treüted mice died from b n e marrow aplasia Myelasuppisession ai 20 and 50 pCi doses with lethality at doscs >50 pCi 30-8574) depletion of bone murow and spl enic hypplasia ND

Sevcre lcukopenia and lethal i iy

Badger et a!., 1986.

Shürkey et al., 1988

Estcban ct al., 1990

Buras et al., 19Wü. Lce et al., 1 990

Abbreviations: 1311, iodine 13 1 ; yttrium 90; iv, intra\!cnous; ip, intraperitonwl; CEA, carcinoein bryonic antigcn; ND, not deicrmined.

RADIOIMMUNOTHERAPY CONCEPTS

Factors lnfluencing Tumour Uptake of Antibodies

Site of Mtninistrarion

The site of administration may be important in the targeting of antibodies to tumours.

Most applications have been by the intravenous (iv) route, which 1s ideal for targeting

hematological or blood borne cancers. However, for compartmentalized cancers such as,

ovarian cancer and colorecial cancer, intraperitoneal (ip) administration may be the most

appropriate for it provides more direct contact between the anlbodies and the cancer cells.

Usi ng B72.3 mA b labeled wi th 13 1 1. researchers have s hown that i p administration to patients

wi th peri toneal carci nomatosis cm achiew improved tumour targeti ng as compared to i Y

administration (Colcher et al., 1987: b o n et al.. 1991a). In addition, targeting turnours

residing in the lymphatics may be best achieved by locai inteetitial administration (Goldenberg,

199 1).

Antibodv Propulies

The antibodies primarily used in RIT are murine IgG rnAbs, which have a molecular

iveipht 01- 150 kilodaltons (ma), when in the intact form. T h e intact IgG consists OS two

identicai heavy chains that weigh 53 Idla each and two identicai light chains, each weighing 77

kDa (Figure 1). The biological functions of the antibody are determined by the constant

regions, while the specificity for antigens is deterrnined by the variable regions. T h e large

molecular weight of the intact IgG antibody irnpedes i ts estnvasation into tumours and i ts

elimination (rom the blood. However, smaller molecular weight mA b denvati ves such as, Fab.

F(ab')z, and scFv fragments and molecular recognition units (MRUs), have k e n developed

(Figure 7). These srnaller molecular weight derivatives retain the an tigen binding specificity of

the intact IgG mAb, and because of their smaller size, can achieve a more homogeneous

tumour distribution, a faster rate of tumour uptake, and a faster rate of elimination from the

blood and normal tissues (Y okota e t al.. 1997). A disadimtage, however, to using these

smaller molecular weight derivatives is that they a n have reduced binding affini ty and avidi ty

cornpared to the intact rnAb (Milenic et al., 1991 ; Y okota et ai., 1992). This factor, combined

with their rapid blood clearance, causes these mAb derivatives to have both a lower absolute

uptake and a shorter residence time in the tumour cornpared to the intact antibody (Buchsbaurn,

1995a).

Murine mAbs are foreign to the human immune system, and therefore, are

irnmunogenic and capable of triggering HAMA responses in patients. Human anti-mouse

aniibodies responses can be generated ei ther against the constant regions (anti-isotypic

antibodies) or the variable regions (anti-idiotypic antibodies) of murine rnAbs. Patients can

elicit a HAMA response as early as 3 weeks following the administration of a single dose of

rnurine mAb (Pimm et al., 1985). which only becomes problernatic after repeated

administrations of the foreign protein. Following repeated administrations, HAMA form

immune compleses with circulating murine mA bs, w hich are then rapidl y sequestered and

deeraded C by the reticuloendothelial system (RES), ie. liver and spleen. resulting in reduced

antibo@ uptake in the tumour (Pimm et al., 1985). Tosicitp from HAMA is rarely observed,

but with repeated, large doses of murine mAb, the immune complexes forrned could

theoretically lead to semm sickness and end-organ damage (Jurcic et al., 1994). Fab, F(ab')2,

and scFv fragments are less immunogenic than intact murine mAbs since they lack large or

entire portions of the constant regions. the regions primarily responsible for triggering immune

responses. Other mAb variants less immunogenic than the intact form include, intact

humanized and intact ctumenc rnA bs (Figure 2). Humanized mA bs are composed of the human

anti body framework, ont0 which, murine complementari ty determi ning regions (CDRs)

directed against the tumour-associated antigen are grafted. The CDRs are sequences of amino

acids located on the variable regions of IgG antibodies that mediate antigen binding (Figure 1).

Chimeric mAbs. which are less human in nature, consist of a murine variable region fused to a

hurnan constant region. Despi te the reduction or elimination of the rnurine constant regions,

HAMA responses, in the form of anti-idiotypic reactions, against the mAb fragments and the

chimeric and humanized mAbs çan occur.

Antibody preparations with the highest possible irnmunoreactivity are desirable. The

immunoreactivity represents the fraction of the antibody preparation capable of binding to i ts

target antigen. Current hybridoma technology is capable of producing large amounts of pure

and specific mAb with LOO% immunoreactivity (Stigbrand et al., 1996). According to results

from iri vin0 studies and experirnentai systems, higher mAb binding affinity correlates directly

with higher tumour uptake (Colcher et al.. 1988). In contrast. some human studies (Gallinger

et al., 1993; Divgi et al., 1994) show no such advan~ge to using mAbs with higher binding

affi ni ty.

Meraboikm of Antibodies

Antibodies are catabolized in lysosomes following intemalization by cells expressing Fc

receptors (Wagle et al., 1978) or cells expressing the target antigen (Press et al.. 1996).

Binding of the antibody to Fc receptors via its Fc region is required for intemalization by the Fc

receptor-bearing cells. The prima- si te of catabolism of anti bodies and other large

macromolecules is the liver, where intemalization is largely mediated by Fc receptors on

Kupffer cells and hepatocytes. Similarly, the reticuloendothelial cells of the spleen and other Fc

receptor- bearing cells also intemalize and catabolize antibodies. Reducing the non-specilic

uptake of antibodies by Fc receptors a n enhance the amount of administered antibody available

for turnour localization. which in tum may lead to increased tumour uptake. In some instances,

administration of unlabeled antibodies pnor to the administration of the radiolabeled mAb may

block the Fc receptors (Perkins and Pirnm. 1991).

Twnoiir Plzysio fogy

The physiology of solid tumours imposes a vanety of barriers. on a micro- and

macroscopic level, that prevent anti body uptake in tumours from reaching optimum Ievels and

achieving hornogeneous distribution. Turnour characteristics such as, vascularity and vascular

permeabili ty, can limi t antibody uptake to oniy well-perfused areas of a turnour. Al though

many tumours may have increased vascular permeability compared to normal tissues, areas of

hi@ interstitial pressure within the turnour can oppose the influx of antibodies (Shockley et al.,

19=). Tumours exhibit a decrease in perfusion with increasing size (Dvorak et al., 1953 1),

and therefore, demonstrate an inverse relationship between an tibod y uptakc and tumour si ze

(Hagan et al., 1986; Williams et al., 1988). In a study by Colcher et ai. ( 1988), up to 3-fold

more specific antibodv localized in LS 174T colorecial xenografts weighing 40- 100 mg, grown

in athymic mice, compared to similar tumour x e n o p f t s weighing 400-700 mg.

Not al1 tumour cells in the tumour express the tumour-associated antigen, which in

turn, l a d s to heterogeneous distribution of antibody throughout the tumour. Higher tumour

antigen dcnsi tv can resui t in greater accretion of antibody, however, tumour localization has

k e n achieved even when on1 y 158 of the tumour ceils espress the mget antigen (Dom et al..

1990). Shockley e t al. ( 1991b) demonstrated in a human melanoma senograft mode1 that

turnour uptake of antibody correlates directly with the level of antigen expression in the

tumour. Moreover, in certain tumours, up-regulation of antigen espression induced by

biological response modifiers such as, interferon, correlated wi th increased anti bod y uptake

(Greiner et al., 1984). However, others have also suggested that targeting highl y antigen-rich

turnours with high affinity mAbs can result in an even more heterogeneous antibody

distribution throughout the tumour. I t is hypothesized that in such tumours, high alfinity mAbs

that extravasate from the tumour capillaries bind primanly to the antigen-positive cells directly

adjacent to the \-asculaturc, creating a "binding site bamer". The "binding site barrier" is not a

physicai barrier caused by the formation of mAb-antigen complexes adjacent to the tumour

vasculature. Rather, the "binding site bamer" is a phenornenon, in which the formation OS

mAb-antigen complexes decreases the nurnber of free mAb molecules that can diffuse into the

deeper regions of the tumour. Hence, antibody distribution in the tumour becomes

heterogeneous, with pronounced binding occumng only adjacent to the tumour vasculature

(Fujimori et al., 1989; Juweid et al., 199-). Antigen shedding from the surface of tumour cells

into the circulation can also advenely affect tumour targeting. Shedding of tumour antigen into

the circulation c m lcad to the formation of mAb-antigen immune complexes in the vascular

cornpartment and subsequently, which can subsequently undergo rapid sequestration and

degradation by the RES (Ullén et ai., 1995).

Considerations for Radioimmunotherapy

In the preceding sections, various biological and non-biological variables that c m affect

antibody uptake in tumours were discussed. These same variables are applicable to

r-dioimmunoimaging and RIT. In order to highlight the complexïties of RIT, the importance of

these and other variables on RIT outcornes are discussed in the following sections.

Atztibodv Properties

Radioimrnunotherapy is a form of systemic therapy that inevitably results in somc

radiation burden to normal tissues. Since intact 1 gG mA bs display very slow el imi nation from

the blood (a-phase half-life of 1 1-15 hours; p-phase hall-life of 18-44 hours) (Perkins and

Rmm. 1991). non-tumour tissues cm be enposed to high doses of radiation for prolonged

periods of tirne as radiolabeled mAb remains in the blood and normal tissues. However,

administration of smaller molecular weight mA b variants such as, Fab, F(abl)*, and scFv

fragments and MRUs, can decrease the ndiation burden to normal tissues because of their

np id elimination from blood (Stigbrand et al., 1996). Of the mAb fragments, scFv Ilragments

(25 ma) display the fastest elimination from blood, followed by Fab fragments (50 D a ) , and

then by F(a& fragments ( 1 0 kDa) (Stigbrand et al.. 19%; Henry et al., 1992). Fab and

scFv fragments have also demonstrated high renal uptake via reabsorption through the

prosimal tubules (Behr et al., 1995; Kobayashi et al.. 1996; Arend and Silverblatt, 1975). and

therefore, can rem1 t in high radiation doses to the kidneys. Renal uptake of these fragments,

however, on be significantly reduced through pre-administration of renal upiake blockers such

as, L-lysine (Kobayashi e t al., 1996; Pimm, 1995). Apart from the improvcd

pharmacokinetics, mA b fragments rnay also offer improved biodistri bution since the lack of a

constant region would reduce Fc receptor-mediated uptake by the liver and spleen. In a

comparative study by Stein et ai. ( 1994), liver and spleen uptake of intact 1311-RS7 mAb in

tumour-bearing mice k v a s at least 5- to 6-fold greater than that of 13 1-RS7 F(abf)2. As a result,

the radiation doses delivered to the liver and spleen were L 9- and 18-fold higher, respectively.

for intact '311-RS7 mAb. In contnst , Lane e t al. (1994) did not find a significant difference

between the liver uptake of intact anti-CEA l 3 11-ASB7 mAb and that of 131I-A5B7 F(abt)2.

Since mAb fragments can display shorter residence times and lower absolute uptake in

the tumour compared to intact mAbs, the radiation dose delivered to the tumour by mAb

fragments could correspondingly, be lower as well. Wahl et al. (1983) compared the uptake of

intact anti-CEA mAb to that of anti-CEA Fab and F(abf)* fragments in tumour xenografts

grown in athymic mice. The authors observed the tumour uptake of intact mA b t o be 2.4-lold

and 76-fold higher than rhat of the F(ab1)2 and Fab fragments, respectively, a t 48 hours

following anti body administration. Similar results were obtained by Stein et al. ( 1994). who

observed an approximatel y 7-fold greater uptake of intact RS7 mA b compared to RS7 F(abl):

in tumour senogd t s , at 24 hours. RS7 F(ab')z fragments, however, displayed a much laster

elimination from blood than intact RS7 mAb. Differences in absolute tumour u p a e and in

climination from blood between RS7 mAb and RS7 F(ab1)2 corresponded wi th similar

differences in the radiation doses delivered to these tissues. At equivalent amounts ol' activity

administered (25 p.Ci), l3lI-RS7 mAb delivered radiation doses to the tumour and blood that

were 4.4- and 70.8-fold higher, respectively, compared to 1311-RS7 F(ab1)2. Despite having ü

lower uptake in turnours, F(ab')2 fragments are cxpected to demonstrate superior anti-tumour

effects and less toxicity compared to intact mAbs, because of their faster rate of uptakc in

tumour, thei r more homogeneous tumour distri but ion, and their shorter biological ha1 f-1 i le.

While only 350 pCi of 1311-US7 mAb was sufficient to achieve a 5 week suppression in

tumour growth, as much as 1 mCi of 1311-RS7 F(ab')2 was required to achievc a comparable

response. At equitosic doses, however. 1311-RS7 F(ab')2 was espected to yield greatcr

therapeutic efficacy than 1311-RS7 mAb. Escalating the dose of 131I-RS7 F(ab')2 by 50% to

1.5 mCi produced an 8 week suppression in tumour growth without any lethality to the

animals. On the other hand, a sirnilar dose escaiation of 50% of '311-RS7 mA b to 375 pCi,

resulted in death of d l the animals.

In cancer patients undergoing estemal radiation therapy (XRT), most normal tissues

can tolerate higher cumulative doses of extemal beam radiation when they are given as several

smaller fractions that are separated by sufficient time for repair, compared to one single dose

(Nias, 1990). Moreover, fractionated RIT has been reported to be more efficacious and l e s

toxic than the traditional single dose RIT. Animal models of fractionated RIT have

demonstrated that larger, fractionated doses of * '1-Iabeled antibody achieve a greater depree of

tumour regression compared to the administration of single rna~imall y tolerated doses (Schlom

et al., 1990; Buchsbaum e t ai., 1990). Schlom et al. (1990) reported that 609 of tumour-

bearing mice died when treated with a single dose of 600 pCi of 1311-mAb. whereas 2

injections of 300 pCi a week apart resulted in death in only 10% of the animals. Fractionatcd

RIT in patients, however, is limited by the HAMA response, particularly ivhen intact munne

mAbs are administered. Despite their decreased immunogeniciiy, mAb fragments and chimeric

and humanized mAbs may not necessarily be more suitable tor fractionated RiT than intact

murine mAbs since these variants are still capable of inducing anti-idiotypic HAMA responses.

In a dose fnctionation study with colorectal cancer patients. 9/12 patients treated with chimeric

1311-B72.3 mAb developed an antibody response to the chirneric antibody (Meredith et al..

199%). However, i t was not determined whether the anti body response was an anti-idiotype

reaction. Two of those nine patients produced such high amounts of antibody againsi the

chimeric B72.3 mAb, that rapid clearance and escretion of chimeric B72.3 mAb occurred,

consequendy resulting in a 6-fold reduction in the whole body radiation dose received from a

second course of treatment. In the study, no correlation between tumour response and antibody

response was observed. Increased radiation doses [O the liver and spleen can result from the

formation of immune complexes with HAMA. On the other hand, HAMA may also protect

patients from bone rnarrow toxicity by reducing the radiation dose delivered to the blood

(Stewart et al., 1989).

Interestingly, anti-idiotype HAMA responses may actually produce an anti-turnour

cffect in some cases. Anti-idiotype anti bodies can mimic the original turnour-associated

antigen, thereby potentially acting as a vaccine and immunizing the patient against the cancer

associated with that antigen. In a clinical radioimmunoimaging trial, 7/32 patients that ivere

repeatedl y administered doses of munne mA b developed a strong anti-idiotypic HA MA

response and surprisingl y, displayed complete rernission or stable disease for a p e n d of 4-42

months (Baum et al., 1994).

Radioisotopes sui table for RIT are listed in Table 3. 1 t is very important to match the

physical half-life oi the radionuclide to the il1 vivo pharmacokinetics of the anti body to help

cnsure that sufficient doses and dose rates for ceil sterilization are delivered to the tumour. If

the half-life of the radionuclide is too short. then most of the decay would have occurred before

tumour uptah oi the antibody reaches masimum levels. Moreover, at quai raùimctivity

concentrations in the tumour, radionuclides with a long half-life will deliver a Iower absorbed

dose rate than those with a shorter half-life (Mausner and Srivastava, 1993). I t has becn

suggested that using a radionuclide with a half-life that is 1- to 3-fold greater than thc tumour

residence time of the antibody is likely to be more therapeutically effective than a radionuclide

rvith either a longer or a shorter hall-life (Rao and Howell, 1993). Altematively, a physical

hall-life similar to the biologcal half-lire of the antibody in the tumour hm also ken suggested

to be ideal (Schubiger and Smith, 1995).

The size of the tumour king treated, ie. solid tumours or micrometastatic disease, and

the dynamics of the target antigen will play a role in selecting the appropriate radionuclide

needed for RIT. These factors will detemine which type of decay: beta (p); alpha (a); or

electron capture; is most suitable for the particular treatment (Table 3). Iodine 13 1, which

Table 3. Radionuclides sui table Ior radioin~miinothewpy.

Radionuclide Decay Mode Physieal t 1 2 Maximum Particulate Energy Maximuni Range of (% Abundance) Particulate Energy In

Tissue (mm) 1311 Beh, gümma 8.0 days 807 kcV ( I)t, 606 keV (86)t, 2.4

336 kcV ( 13)t 6 7 C ~ Bciü, gamma 67 heurs 577 kcV (70)t, 484 kcV (35)t 3 -.- 3

395 keV (45) f 90Y Beta 64 hours 329 MeV ( 10O)t 11.9

1 8 8 R ~ Bctü 17 hours 2.13 MeV ( I00)t 1 1 . 1 212Bi Alpha, bclü 1 hour 6.09 MeV ( IO)#, 6.05 MeV (25)* , O. O9

5.77 MeV ( 1 )+, 2.27 MeV (40)t, 1.55 McV (7)t, 930 kcV (5)t, 670 ke V (4) t ,450 kc V (5) t , 85 kcV (3)t

211At Alpha 7 hours 5.87 MeV ( 100)~ 0.09 1251 El cc tron capt urc 60 driys 35 kcV ( 100) 0.02

t Beta irradiation. t Alpha irridiiition. Abbreviütions: lwI, i d n c 13 1 ; b 7 C ~ , wpper 67; POY , yttrium 90; lmRe, rheniuni 188; * Wi. bismuth 11 2; 21 ]Ai, miatinc 2 1 I ; 1251, iodinc 135.

decays by both gamma (y ) - and $-emissions is an atiractive candidate for RIT. The y-emissions

do not deliver sufficient radiation doses to cause a tumouncidai effect, but instead, allow for

the use of extemal scintigraphie irnaging to determine the biodistribution of the radiolabeled

antibody and to estimate the absorbed radiation doses lrom the fLemissions delivered to the

tumour and normal tissues. However, the y-emissions c m deliver unnecessary radiation doses

to normal tissues due to their long emission ange.

Beta-emitters such as, yttrium 90 o r rhenium 188 (IgWe), can deposit their

particulate energy up to a maximum range of 1 1.9 and 1 1.1 mm in tissue, respectively, makrng

them suitable for treating relatively large solid tumours. For tumours d mm, however, copper

67 (67C~) and 1 3 1 1 . which possess maximum f.3-emission ranges of 1.5 and 1.4 mm,

respectively. would be preferable over 90Y and 18sRe because of their shorter range of

emission. Using these radionuclides increases the likelihood that most of the energy released

iorm the p-emissions is deposited within the tumour and not in neighbouring normal tissues.

Mathematical modeling suggests that the optimum tumour diameters for masimizing the

radiation dose delivered to the lesion (and potentially cunbility) are 28-42. mm, 13-32. mm,

3.6-5 mm, and 1.6-7.8 mm when 90Y. lggRe, 1311, and 6 7 C ~ , respectively, are utilized

(O'Donoghue, 1996). The model also suggests that turnour curability could potentially

dccrease when the tumour diameter fails above or beloa these optimum ranges. I î localized in a

tumour with a diameter well below the calculated optimum range, would deposit rnost of

i t s energy outside the lesion and in adjacent normal tissues. The model, however, assumes

uniform tumour distribution of radionuclide and an equivalent phenotype, proliferation rate,

and radiosensi tivi ty amongs t the tumour cells.

Because of their relatively long range, fkmitters such as, l 3 11, o r 67Cu, can

produce a "cross-fire" effect (Nourigat et al., 1990). When bound to tumour cells, antibodies

Iabeled with these radionuclides c m irradiate nearby tumour cells that were not targeted by the

radioantibody. Thus, heterogeneous antibody distribution in the tumour, arising from

physiological barriers, ie. the "binding-site barrier" phenornenon, poor vasculanzation, or

heterogeneous antigen distribution, rnay not necessarily hinder their tumouricidal effect. In

contrast, using antibodies labeled with a-emitting radionuclides such as, bismuth 111 (2 12Bi j

and astatine 2 1 1 (21 Nt), may be ineffective when heterogeneous antibody distribution in the

tumour occun, due to their estremely short range of 4 Pm, ie. 4 0 ce11 diameters. More

suitable tumour targets for a-emitters include, lymphoma, micrometastases, and compart-

menrally disseminated neoplasms such as, neoplastic meningitis (Zalutsh~ and Bigner, 1996).

Radionuclides that decay by electron capture, îe. iodine 175 (12j I ) , have the shortest range oof

energy deposition, typically -0 yrn. Thus, their ability for efficient tumour cell kill ivould

require intracellular deposition, ie. intemalization of the antigen-antibody cornplex. Due to their

short ernission ranges, 212Bi. 21 IAt, and 1251 lack a "cross-Tire" effect but, on the other hand,

can spare normal tissues from unnecessaq irradiation.

lodine 13 1 and g O Y are the two most commonl y used radionuclides in esperimental and

clinical RIT studies. M i n e 13 1 has several advantages including, well-known ndiochemistry,

ready availability, and low cost (Wilder et al., 1996). On the other hand, the emission

properties of 9 0 Y enables it to deliver approximately 4.5-fold more radiation to a tumour than

1, per mCi (Larson, 199 1 b). However, this also means that 90Y c m deliver more radiation

than l 3 I 1 to normal tissues. In patients treated with I 3 '1- and 90Y-rnAbs, Stewart et al. ( 1988)

observed that was 8-fold more tosic to bone marrow, per mCi, than 13 '1. These authors

proposed that bone absorption of free can be the dominant source of bone mmow

irradiation and toxici ty from this radionuclide. On the other hand, Vriesendorp et al. ( 1996)

suggest that circulating 90Y, labeled to a high molecular weight, slowly eliminated molecule,

ie. an intact IgG mAb, is the dominant source of bone marrow irradiation and tosicity.

Nevertheless, both 90Y and 1311 have shown some success in clinical RIT. In a Phase 1/11

study by Lane et al. ( L994), ten patients wi th colorectai cancer received anti-CEA l 3 lI-A5B7

mAb (36-64 mCi), while nine patients received I3'1-A5B7 F(ab')z (82-148 mCi). One

complete response was observed in the group that received l3 lI-A5B7 and one partial response

was observed in the group treated with 1311-A5B7 F(ab')z. In both groups, however,

myelotoxicity was the limiting factor. In another Phase 1 triai, al1 of twenty-four colorectal

cancer patients treated with 1311-CC49 (45-128 mCi) showed no major responses but, six

patients did dernonstrate stable disease for 4 weeks following initiai treatment. Myelotoxicity

was observed in eleven patients (Divgi et al., 1995). Knox et ai. ( 199%) treated eighteen

patients with recurrent B-ce11 non-Hodgkin's lymphoma with 1-2 intravenous doses of anti-

CD20 9 V - 2 B 8 mAb ( 13.5-50 d i ) . Four patients demonstrated complete responses. ninc

patients showed partial responses but, myelotoxicity was observed.

Follow ing cellular intemaiization, ndioiodinated anti bodies rapidl y undergo proteolpsis

and deiodination by degradative enzymes including, dehalogenases present in the lysosomes ol'

tumour and certain normal tissue cells, ie. phagocytic cells such as, blood mononuclear cells

and reticuloendothelial cells, resulting in the cleavage of radioiodine from the antibody (Wilder

et al., 1996). Following lysosomal degradation, free radioiodine and small iodinated peptides

are released from the cells and subsequently, are elirninated from the body by rend excretion,

or in the case of free radioiodine, are secreted into the stomach or sequestered by the thyroid or

salivary glands. On the other hand, cortical bone and liver are the major deposition sites for

free (Durbin, 1x0). The primary route of elirnination of free 90Y is rend excretion,

however, some of the deposited in the liver map be escreted through the gastrointestinal

tract (Durbin, 1960). Compared to ndioiodinated an ti bodies, g o Y -1abeled an tibodies display

superior cellular retention cvhen intemalized and catabdized by tumour celk (Press et al.,

19%) or normal blood mononuclear cells (Nanilri et al., 1990). Radioiodine is conjugated to

anti bodies pnmarily by substitution ont0 tyrosine arnino acids, whereas 90Y is chelated ont0

anti bodies through chelating agents such as, diethylenetriarninepentaacetic acid (DTPA).

Chelating agents such as, DTPA, cm be conjugated ont0 the antibody via lysine residues

(Hnatowich and McGann, l m . Yttrium 90 ions circulating in blood can also be scavengeà

by transfemn (Moi et al., 1990).

Radio biology

It has been suggested that to treat cancer successfulIy, a total radiation dose of 50-80

Gy must be delivered to the tumour (Bruland, 1995). The induction of double-strand breaks in

the DNA is ûelieved to be the mechanism by which RIT eiicits tumour ce11 hll, rendering

tumour cells incapable of division and growth (Cobb and Humm, 1986). Linear tnnsfer

energy represents the average energy l d l y irnparted to a medium by a radiation particle ivhen

traversing 1 pm dong its path (Hall, 1994). Since fbparticles deposit their energy over a longer

range (mm) compared to a-particles (pm), they are low linear transfer energy radionuclides,

wherûas a-emi tters are high linear transfer energy radionuclides.

High linear transfer energy radiation is densely ionizing and is more efficient at

producing DNA breaks cornpared to low linear transfer energy radiation. As a result, DNA

breaks produced by a-particles are produced directly and are irreparable. In contmt, the DNA

breaks produced by fbparticles are done so indirectly, via highly reactive molecules, and are

susceptible to repair (Hall, 1994). When $-emitting radionuclides are used, approsimatel y 200

DNA double-strand breaks per ce11 are required to sterilize 99% of a tumour ce1 l population,

whereas only a few DNA breaks are required when a-emitters are used (Cobb and Humm,

1986). Similarly, only 3-6 impacts per nucleus by a-particles are required to sterilize up to

67% of tumour cclls, but for fi-particles, up to 400 times as many impacts are requircd

(Humm, 1986).

In terms of dose rates, a-particles are cytotosic at rates as low as 0.01 Gylhour,

whereas, even at dose rates of 0.1 G-hour, p-particles may not be tumouncidal (Fowler,

1990). Moreover, since DNA breaks produced by high linear transfer energy radiation are

irreparable, a-emitters are cytotoxic even to hypoxic cells (Brown, 1986). For DNA breaks

produced by p-emitters, however, hypoxia significantly reduces cytotoxicity (Hall, 1994).

Since tumours are known to contain viable hypoxic cells, such radiobiotogical diffcrences

between a- and fi-emitters can be signifiant.

The efficiency of RIT will depend on the total radiation dose delivered to the tumour,

the dose rate, the duration of tumour irradiation, and the rate of cell proliferation. Andres et d.

( 1986) suggest that a minimal dose rate of 0.4 Gylhour must be delivered to the tumour over a

long enough period of Ume. so that a i l tumour cells are imdiated dunng the rnost radiosensitive

M and G2 phases of their ce11 cycie. Typically, conventional XRT delivers dose rates of 60-

180 Gylhour to the tumour, while RIT provides dose rates of 4 . 6 Gylhour (Hall, 1988). The

dose rate delivered by XRT is constant and is delivered to a limi ted region of the body, ie. the

tumour and portions of adjacent tissues. In contrast, the dose nte delivered by RIT initially

increases as the radioanti body accumulates in the tumour. then continuousl y decreases due to

biological clearance and physid decay (ODonophue, 1994). Moreover, RIT delivers radiation

to the entire body, not just the tumour and adjacent tissues. Labeling an antibody with a

radionuclide such as, which ernits more energy per decay and has a shorter physicd half-

life (2.7 days). instead of l 3 l I (haif-Me of 8 days). can produce a dose rate up to IO-fold

greater ( Wessels and Rogus. 1984).

Results from esperimental tumour models have ken quite variable in that the relative

biological effectiveness of' RIT has shown to be less (Buchsbaum et al., 1990), equal (Buras et

al ., 1 WOb), or supenor (Wessels et al., 1989) to that of XRT. Possible esplmations for the

obscrved superior anti-tumour effect of RIT compared to dose-equivalent XRT. ivhich is also

known as the "inverse dose rate effect". include the following (Knos. 1995b): i) RIT targets a

rapidly proliferating subpopuiation of tumour cells that are primarily responsi blc for tumour

doubling (Wessels et al., 1989); i i ) rcoxygenation of hyposic tumour cells during the

protracted course of RIT increases their radiosensi tivi ty (Hall, 1994) ; iii) loir. dose rate

irradiation induces apoptosis to a larger degree than high dose rate irradiation (Macklis et al.,

1993); iv) the biological effects of some antibodies increase tumour blood flow and elici t an

inflammatory response (Sands et al., 1988); and v) low dose rate irradiation leads to the

accumulationof tumour cells in the radiosensitive G2 phase of the ce11 cycle (Knos et al.,

1993).

T o date, RIT of leukemias, lymphomas, and neuroblastomas has k e n most effective,

whereas RIT of most solid tumours has been less promising (Kairerno. 1996).

Radioimmunotherapy of lymphomas, in particular. has achieved the best clinical results and

offers great prospects for success. As much as 50% of patients with non-Hodgkin's B-ceil

lymphoma have achieved complete responses from RIT (Wilder e t ai., 1996). Currently,

leukemias and lymphomas appear to the most attractive candidates for RIT for several reasons:

i) the cancers are relatively radioresponsive (Lybeert et al., 1987); i i ) patients often have

diminished HAMA responses due to supressed humoral immunity. which allows dose

fractionation (DeNardo et al.. 1995); and iii) leukemias and l ymphomas, ie. hematological

malignancies, are more readily accessible to antibodies than cells of solid, bulky tumours

(DeNardo et al., 1988)-

Limitations of Radioimmunotherapy

Response rates in clinical RIT studies (Table 3) have yet to match those obtained from

animal studies (Table 1). Extrapolating the success from animal studies to clinical trials is

difiÏcuIt because of the inherent differences between animal models and humans. These include

the following (Buchsbaum, 1995b; Knox, 1995b; Schubiger and Srni th, 1995): i) di fferences

between the phamacokinetics and biodistribution, ie. tumour uptake in animai models is 10-

35% of the injected dose per gram of tissue (B id/@, whercas t0.0 1%, idlg CS typically

achieved in humans; ii) differences in the volume of distribution and the nurnber of tirnes the

ndiolabeled anti body circulates through the tumour vasculature; iii) the lack of HAMA

responses in the murine model; iv) higher radiosensitivity of human bone marrow cornpared to

murine bone marrow (Badger et al., 1985); and v) tumour xeno@ts in the animal model often

have a better blood supply compared to tumours in patients (Goodwin, 1987).

Meanwhile, radioirnrnunoimaging has enjoyed greater clinical success, as is evident

horn the approval of two cancer imaging agents, Oncoscint CR/OV@ and CEA-SC~~@, by the

US FDA for clinical use. Despite the tremendous amount of intense research dedicated to NT

during the past 20 years, the major limitations of RIT have yet to be resolved, which include

the following: i ) only a minute fraction of the antibody dose localizes in solid tumours

( 4 . 1 % ) . resulting in the de l ives of sub-therapeutic doses (4 Gy); ii) the dose-limiting bone

marrow toxicity, which has k e n observed in virtudly ail clinicd studies; and iii) the HAMA

response, which lirnits repeated courses of treatment as in dose fractionated RIT.

Strategies to improve Radioimmunotherapy

In order to improve the clinical success of RIT. various approaches to improve

antibody uptake in tumours, enhance tumour cytotoxicity, and reduce bone marrow toxicity

and HAMA responses, have k e n investigated.

Impro ving Twnorrr Uptake of An fibodies

As described previously, using mAbs with greater affinity may increase their uptake

and retention in turnours. Colcher et al. ( 1988) compared the turnour u p a e of 872.3 mAb and

its second generation rnAb, CC49. in LS 174T senografts grown in athyrnic mice. CC49 mAb,

which has a binding affinity constant (Kk) 6-fold greater than that of B77.3 demonstrated an

approsimately Xold higher uptake in the tumour. When studied in colorectal cancer patients,

however, B72.3 and CC49 mAb did not show any significant differences in tumour uptake

(Gallinger et al., 1993; Divgi et al., 1994). This discrepancy couid be due to the presence of

significant amounts of circulating target antigen, which favours antibody binding to the

shedded an t ign nther than the turnour-associated antigen on the surface of tumour cells or, to

the overriding influence of the many barriers that tumour physiology imposes on antibody

uptake.

Researchers have aiso attempted to increase tumour expression of the target antigen in

order to increase tumour uptake of antibodies. Most of the research in this area has focused on

the CEA and tumour-associated glycoprotein 72 (TAG-72) antigens. Experimental, as well as

clinicai evidence, have shown that interferon treatment c m up-reguiate the tumour surfacc

expression of CEA and TAG-73 antigens in a variety of cancer cells (Kantor et al., 1989;

Greiner et al., 1990; Rosenblum et ai., 1988; Murray et al., 1995). Elevation of specific

mRNA transcripts following interferon treatment suggests that upregulation is a result of new

antigen synthesis (Kantor et al., 1989). The abili ty of interferon to enhance the therapeutic

efficacy or RIT ivas demonstrated in an animal mode1 (Greiner et al., 1993). in which

administration of the cytokine to athymic mice bearing HT-39 colon cancer senografts

increased TAG-77 levels L O-fold and correspondi ngl y, increased tumour localization of anti-

TAG-72 13 ILCC49 mAb 3- to Cfold. Mice that were administered interferon-y demonstrated a

better thenpeutic response to treatment with 300 pCi 13WCC49 compared to those not

receiving interferon-y. In a clinical trial, patients wiih late stage breast cancer were administered

a-interféron in corn bination with therapeutic doses of 13 11-CC49 mA b (Murray et ai ., 1995).

Some clinical responses were observed, however, the interferon and l 3 ' 1-CC49 mA b

corn bination may have aiso enhanced marrow toxicity.

Increasing intratumoural vascular permeabili ty rnay also be effective in increasing

turnour uptake of antibodies. Experimental evidence has shown that interleukin-2 (IL-?),

administered in its free f o m (DeNardo et al.. 199 1) or conjugated to the targeting antibody

(LeBerthon et al., 1991), can increase vascular permeability and antibody uptake in tumours.

Increased vascular permeabili ty may be a result of the interaction between activated

tymphocytes and endothelid cells or through some direct effect of IL-? and the endothelium

(Allison et ai., 1989). Low-dose XRT of the turnour has also k e n shown to be effective in

increasing intratumoural vascular permeability in animai modeis and in patients. A total dose of

6-9 Gy from 2-3 courses of XRT ending one day prior to antibody administration achieved an

approaimately 3-lold improvement in tumour uptake in patients with hepatoma (Msirikale et

ai., 1987). Contnsting experimental evidence, however, suggest that this phenomenon may

not occur with al1 tumours (Shnvastava, et al., 1989).

Other methods under investigation include, the use of local hyperthermia to reduce

interstitiai fluid pressure and enhance antibody uptake (Wilder et al., 1993; Leunig et al.,

1 992), administration of antibody "cocktailsn, which are composed of anti bodies specific for

different epitopes on the tumour cells (Blumenthal et al., 1991), and "pre-dosing" with

unlabeled a n t h d i e s . 'Re-dosing" attempts to saturate Fc receptors on cells responsible for

mAb catabolism and to form immune complexes with circulating antigen, prior to administerîng

the ndiolabeled antibody (Buchsbaum et al., 1992).

Erthnncing Tumocir îytotoxicip

Animal models have shown that the proportion of hyposic cells in the turnour increases

as the turnour grows from rnicroscopic to macroscopic size (Moulder and Rockwell, 1984). In

some patients, tumour hyposia has correlated with reduced efficacy of XRT ( H k k e l et al..

1993; Okunieff et ai.. 1993). Tumour hyposia could also potentially diminish the elfectiveness

of RIT, particulad y when loir Iinear transfer energy radionuclides are employed. 1 rnproved

tumour responses to radiation therapy have been reported following the administration of

hyperbaric osygen (Henk and Smith, 1977) or hyposic ce11 sensitizers (Urtasun et al., 1976).

Improved tumour responses to RIT following the use of phannacologicai agents such as,

tirapazamine (previously known as S R 4333). a direct-acting, first generation hyposic

cytotosin, have been demonstrated in animal models (Wilder et al., 1993 ; 1994). Langmuir and

Mcndonca ( 1992) dso demonstrated that RIT corn bined wi th SR 4233 treatment u-as at least

10-fold more cytotosic than radiolabeled antibody alone. Esperirnental evidencc suggests rhat

under hyposic conditions, tirapazamine becomes a free radical and is capable of inducing DNA

strand breaks, leading to ce11 death during mitosis ( B r o w and Lemmon, 1990). In addition to

increasing intratumoural vascular permeability. local hyperthennia can also be directly cytotosic

to certain ndioresistant tumour cells, ie. hyposic cells, acutely acidic cells, and cells in the S

phase of the ceil cycle (Hall, 1994; Sneed and Phillips, 1991). Both the temperature and

duration of the treatment determines the degree of ce11 blling achieved. In a study by Wilder ct

al. ( l m ) , local hyperthermia treatment for 1 hour and at a temperature of 43'C sigmficantly

enhanced the effectiveness of RIT of athymic mice beanng colorectai senografts. I n the same

study, a greater anti-tumour effect aas observed when local hyperthermia and tirapaismine

were combined with RIT.

Reducing HAMA

Attempts to abrogate HAMA have largely focused on the development and use of less

immunogenic forms of rnA bs such as, Fab. F(ab')2, and scFv fragments and hurnanized and

chirneric intact mAbs (Figure 2). Alrhough these derivatives lack large o r entire portions of the

murine constant regions, the possi bili ty of HAMA responses directed against the variable

regions eBsts (see: "Factors Influencing Tumour Uptake of Antibodies - Antibody Properties"

and "Consideritions for Radioimmunothenpy - Anti body Properties"). Based on the method

for the production of murine mA bs established by Kohler and Milstein ( 1975). researchers

have attempted to produce human antibodies (Glassy e t al., 1987: Kozbor and Roder, 1981 ).

These would also be much less immunogenic than intact munne mAbs, however, the instability

of the hybridoma cells and iow mAb yield remain the biggest limitations (Schubiger and Smith,

1 995).

Phmacologica l intervention with immunosuppressants such as, cyclosponn A, has

also been attempted. In one clinical study, thirteen patients were piven oral cyclosponn A two

da' pnor to an initial administration of Fab o r F(abf): fragments (Weiden e t al.. 1994). Sis to

9 days later. patients then received either lg6Re-labeled F(abV)* o r an intact m A b and ucre then

given an additional course of oral cyclosporin A for up to 14 days. Of the patients w-ho

received cyclosponn, signi ficant suppression of HAMA \vas obsen-ed in those uri th relative1 y

hi gher cyclosponn levels. The cyclosponn- treated group also eshi bited si@ fican t HAMA

suppression compared to the patients who did not receive the immunosuppressant.

Rather than suppressing the HAMA response, attempts have been made to reducc the

levels of circulating HAMA. Plasrnapheresis, a procedure in which plasma is separated from

cellular blood components has been s h o w to be effective in reducing HAMA levels in

patients. Patients in a RIT trial for the treatment of cutaneous T-ce11 lymphoma undergoing

plasmapheresis had their HAMA ti ters reduced by 38-6 1% (Zimmer et al., 1988).

Reduciug Bme Marraw Toxici',

Currentiy, the dose-lirniting organ of RIT is the bone marrow. GranuIocytopenia and

thrombocytopenia. the two most cornmon forms of myelotoxicity, reach nadir levels at

approsimatel y 4-6 weeks after the start of therapy. In most cases, the myelosupression is

reversible, with white blood ce11 counts reaching normal levels 2-4 weeks later. Other side

effects in other organ systems are rare and occur only after serious dose escalation

(Vriesendorp et al.. 1991) or with combination therapies (Wang et al., 19%). For esample,

after increasing the dose of 9V-anti-femtin mAb 4-fold to 4 mCikg, Vnesendorp et al. (1991)

observed radiation hepatitis and massive ascites formation in beagle dogs. aî 5 months

following therapy. A similar dose escalation lead to the development of gastrointestinal tosicity

in rats 5 days after treatment. In another study. Wang et al. ( 1995) observed radiation hepatitis

and formation of ascites in beagle dogs 30 days after king treated with 90Y -ZCE0?5 mA b (0.5

rnCilkg) and estemal beam radiation of cobalt 60 to whole liver (30 Gy fractionated into 2

G - d q at 5 dayslweek for 3 weeks). The dogs eventually developed terminal liver failure.

Hematopoietic stem ceII darnage seerns to be the most significant mechanism of bone marrow

tosicity. Reducing or abrogating RIT-associated myelotoxicity is important because it could

potentially permit higher amounts of radioactivity to be safely administcred to patients, possibly

improving the radiation dose delivered to the tumour.

Reducing the biological haif-life of the radiolabeled antibody in the blood will reduce

the radiation dose delivered to the bone marrow. Administration of secondq antibodies, ie.

goat anti-mouse anti bodies, as a clearing agent have demons trated some success in reduci ng the

bone marrow dose to patients (Begent et al.. 1987). Secondary antibodies are directed against

the radiolabeled antibodies and together form immune complexes in the circulation. The

immune compleses are then quicMy removed from blood in a manner analogous to that oi the

immune complexes fomed by HAMA. Another approach to enhance the clearance of the

radiolabeled antibody involves the use of biotinylated anti-tumour antibodies. folloaed by

administration of avidin or streptavidin as clearing agents. The very high affinity of the

(strept)avidin/biotin interaction promotes the formation of immune complexes in the circulation.

which are also quickly removed from blood. After administering streptavidin (at a IO-fold

molar escess of the administered dose of radioamibody) to tumour-beanng mice 74 hours &ter

administration of biotinylated -anti-CEA mAb, Marshall et ai. ( 1991) obsen-ed a 40%

reduction in blood radioacti\.ity levels. Interestingl y, when the number of biotin molecules

conjugated to the radiolabeled antibody n.as increased from four to nine biotins per mAb. blood

radioactivity levels decreased eïen further, by 1CfoId. By companng the anti-tumour effect of

radiolabeled antibody alone vs. that of radiolabelcd antibody followed by second- antibod~,

Biurnenthal et al. ( 1989) obsen-ed that administration of a second- ami body ai lowed a 2-fold

increase in the administered dose of radioactivi tu, resulting in increased efficacy against the

tumour.

Fab, F(ab')-, and scFv fragments or MRUs offer an alternative approach to decrease

the radiation dose to the bone marron.. As described previousl y. these derivatives possess a

îàster b l d and normal tissue elimination compared to intact IgG mAbs as weli as different

properties with regard to tumour uptake and distribution (see: "Factors lnfluencing Tumour

Uptake of Antibodies - Antibody Propertiesn and "Considentions for Radioimmunotherapy -

Antibody Properties" ). Despite promising results in espenmentai systems, so far no signi ficant

improvement in the therapeutic indes over intact IgG mA bs has been obsen-ed in patients

(Stigbrand, 19%; Lane et al.. 1994; Mach et al., 1988). Further evaluation of mAb fragments

wi i i be required to assess their clinicd utili ty in RIT.

Fractionated RIT îs an approach that is less tosic to bone marron-. Meredith et al.

( 1W-b) treated twelve colorectai cancer patients u.i th either 3 s 14 mCi/rnT 2 s 18 mCi/mz, or

3 x I I mCi/mz of 1311-chimeric B71.3 mAb. The fnctionated dosing schedule reduced the

degree of bone marrou tosicity. and produced a moderate increase in the thenpeutic indes.

However, despite using the less irnmunogenic chimeric form of the B72.3 mAb, HAMA

responses were observed in 7 5 4 of the patients. In another study. dose fractionation of

patients with B-ce11 rnalignancies resulted in ten complete or partial remissions in eighteen

patients (DeNardo et al., 1990). The total radioactivity administered (300 mCi) iras C ~ i v e n in

30-60 mCi fractions and resulted in thrombocytopenia in only tuvo patients. Although intact

munne mA b sras utilized, the frequency of HALMA \vas low, possi bl y refiecting the depressed

immune funcrion that is typical of patients with B-ce11 malignancies. Obviousl y, future cli nical

success of fractionated RIT wiil require reduction of HAMA responses. Other attempts to

reduce bone marron. tosicity include relatively more cornplicated procedures such as.

autologous bone marrow transplantation ( ABMT) (Press et al.. 1989; 1993) and extracorporeal

immunoadsorption (ECIA) (DeNardo et al., 1993). In ABMT. the patient's bone marrow is

han-ested pnor to therapy and reinfused aftenrards to restore the bone manou stem cell

population. whereas in ECIA. plasmapheresis techniques are used to remove escess circulating

rad i o m ti bodies.

Adjunctive use of colony stimulating factors (CSFs) such as, interleuhn- l (IL- 1 ), has

k e n used to protect the bone rnarrow from the damaging effects of radiation. Interleuhn- 1

protects the hematopoietic stem cells from radiation (Gallicchio et al., 1989) and stimulates

stem cell grou-th (Neta et al.. 1987). Blumenthal et al. ( 1988) demonstrated in an animal mode1

that IL- 1 treatrnent 20 hours pnor to the administration of W-Iabeled mAb could pre\-ent bone

rnarroir toxicity. In addition, IL- 1 treatment 7 days lollowing therapy Kas capable of re\.ersing

the bone marron- toxicity induced by the 1311-1abe1ed antibod>.. Interleuhn- 1, hou.et-er, is not

capable of stimulating the growth of megal<anocytes and therefore, may have l e s of an effect

on reducing thrombocytopenia.

Multi-step o r pre-taqeted RIT is another approach that can ptentiall- reduce the

radiation dose delivered to the bone marron.. Various pre-targeting protocols currentlv being

inwstigated include. the cornplementa.ry DNA oligonucleotide system (Kuijpen et al., 1993).

the bispecific antibody-bivalent hapten system (Le Doussal et al., 1989). and the (strept)a\-idin-

biotin system (Paganelli e t ai., 1991a). In the latter two approaches, an antibody possessing

binding affinity for both the target antigen and a small molecular weight effector molecule is

administered to the patient. Following sufficient tumour uptake of the bifunctionai antibody and

clearance from the blood and normal tissues, the effector molecule, labeled with a therapeutic

radionuclide. is then administered to deliver the radioactivity to the tumour. Escess effector

molecule not localized in the tumour is then quickly eliminated from the body, ideally by rend

escretion. Too long a delay between administration of the bifunctional antibody and effector

rnolecule may resdt in sub-optimum tumour radiolocaiization due to diffusion of the antibody

from the tumour. The spatial distribution of the effector molecule in the tumour would be

dictated by the intratumourai distri bution of the anti body, and hence, appropriate radionuclides

would have to be chosen on the ba is of tumour size, antigen distribution, and the radiolabeling

methods avaiiable. However, beause of the small size of the effector molecule, diffusion [rom

the vascular cornpartment into the turnour would be considenbly faster than that of intact mAb.

Therefore, the effector molecule could potentially deliver a higher initial dose rate to the

tumour, assuming that similar radionuclides are used and absolute tumour uptake is not

reduced due to rapid elimination of the effector molecule. Antigen intemali-zation could have a

profound effect on the effectiveness of pre-targeting protocols since it could reduce the number

of binding sites available for the cffector moIecule on the surface of the tumour cells.

Non-specific irradiation of the bone marrow in pre-targeting approaches should be

relatively small compared to that from RIT with radiolabeled intact mAbs or even F(ab')2

fragments, since the effector molecule bearing the radionuclide is typically 51-3 kDa and

exhibits a faster elimination from the blood. The ideal characteristics of an effector molecule

include: i) low molecular weight and hydrophiiici ty ; ii) rapid elimination from blood and rend

escretion; and iii) little o r no affinity for normal tissues. The half-life of certain effector

molecules in the blood typically lasts for several hours, whereas that of intact murine mAb is

typicall y several days long. Thus. pre- targeting is able to achieve high time-dependent

tumour: normal tissue (TNT) ratios more rapidl y than conventional targeting wi th radiolabeled

mAbs. For example, administration of biotinylated AUA 1 rnAb followed by avidîn 74 hours

later (to clear excess antibody from blood) and then radiolabeled biotin, to tumour-beanng

mice, achieved a tumoun blood (T: BLD) ratio of approximatel y 50: 1 on1 y 2 hours followi ng

injection of radiolabeled biotin. In contrast, administration of 1311-AUA 1 mA b achieved a

T:BLD ntio 4: 1, more than 50-fold less than that achieved by the pre-targeting method

(Paganelli et al., 1990a). Tumour: normal tissue ratios, which provide indices OS

radiolocali7ation. are caiculated as the ratio between the % idg in tumour and the Q idlg in a

normal tissue.

Pre-targeted RIT is an attractive approach to reduce bone mmow tosici ty because: i ) i t

does not require the use OS phmacological adjuvants, ie. IL- 1 ; ii) the elirnination half-li fe of

radioactivity in the blood for pre-targeted RIT is shon compared to that of radiolabeled mAbs

and even F(ab')2 fmgments; and iii) i t avoids cornples medical procedures such as, ABMT or

EClA ,which could be associated with signifiant morbidity and mortali ty nsks. Al though there

are much fewer expenmental and clinicai studies of pre-targeted RIT that have been conducted

compared to conventional EUT, recent results are very encouraging (Axworthy et al., 1995;

Cremonesi et d.. 1997; Le Doussal et al.. 1989). One promising approach for pre-targeted RIT

is that involving the (strept)avidin-biotin system, which has previousl y been investigated by

Our laboratory in pre-targeted radioimmunoimaging of LS 174T human colorectal senografts

grown subcuianeously in athymic mice ( Alvarez-Diez et al., 1996).

PRE-TARGETING WITH THE (STREPT)AmIN-BI= SYSTEM

There are three variations to the (strept)avidin-biotin system for pre-targeted RIT

(Figure 3). These include: i) biotinylated mAb followed by radiolabeled (strept)avidin (7-step);

ii) (strept)avidinylated mAb followed by radiolabeled biotin ( h t e p ) ; and iii) biotinylated mAb

followed by, (strept)avidin as a clearing agent, and radiolabeled biotin (3-step). These

approaches exploit the very high binding affinity between the (strept)avidin and biotin

molecules for tumour r ad io ld i za t ion . Application of the (strept)avidin-biotin system to target

specific antigens is not unique, for i t is also widely used for in v i~ro applications in

immunohistochemistry, enzyme linked immunosorbent assay, and molecular biology (Wilchek

and Bayer, 1988).

Properties of (Streptlavidin and Biotin

Avidin molecules are a Camily of oligomenc proteins that are approsirnately 65 kDa in

weight and are made up of four identical subuni ts, each weighing approsimatel y 16 D a . Each

subunit can bind one molecule of biotin and thus, one molecule of avidin can bind up to four

molecules of biotin. The binding sites, however, are fairly close to each other, which may offer

sorne steric hindrance to the binding of large biotinylated molecules, in which case, the avidin

protein ma): not necessarily behave tetravalently. The binding affinity between avidin and biotin

is 1015 M-1 (Green, 1%3), approsirnately 1 million-fold greater than that of most antibody-

antigen interactions. For practical purposes, the binding of biotin to avidin c m be regarded üs

an irrcversible process. T h e family of' avidins include, protcins that are produced by

arnphibians, reptiles, and avians (Woolley aqd Longsworth, 1942; Korpela et al., 198 1), as

weil as that produced by S~repzomyces avidinii, which is h o w n as streptavidin (Stapel y et al.,

1963). AI1 avidins possess identical biotin-binding properties but, differ in physico-chernical

properties. For enample, avian avidins are very basic glycoproteins with an isoelectric point of

approximately 10.5 (Woolley and Longsworth, 1942), thereby enisting as a cationic species at

physiological pH. Thus, in vivo, they are capable of associating with negatively charged

compounds such as, mucopolysaccharides or nucleic acids. Streptavidin, on the other hand,

1. Biotinylated mAb and radiolabeled (strept)avidin (Estep)

--- --

-P Ir Al ow for tumour uptake - m a and normal tissue elimination

\ Radiolabeled (strept)avidin

Turnour associated antigen

II. (Strept)avidinylated mAb and radiolabeled biotin

n R *

Mow for tumour uptake and normal tissue elimination

III. Biotinylated mAb, (strept)avidin, and radiolabeled biotin

.41Iow for tumour uptake and encourage elirnination of biotinylated mAb

Figure 3. Schematic representation of the three variations to the (strept)avidin-biotin pre- targeting system. 1. Biotinylated mAb is first administered to the patients followed by radiolabeled (strept)avidin after sufficient tumour uptake and normal tissue elimination of antibody has occurred; I I . (Strept)avidinylated mAb followed by radiolabeled biotin; I I I . Biotinylated mAb followed by unlabeled (strept)avidin following sufficient tumour uptake of the antibody. (Strept)avidin works as a clearing agent and provides binding sites for radiolabeled biotin in the tumour. After elimination of (strept)avidin-biotinylated mA b complexes, radiolabeled bioti n is adminis tered.

does not contain carbohydrate residues and is uncharged at physiological pH (Dittmer et al.,

1989).

Biotin, Iormerly known as vitamin H, is a 244 Da molecule consisting of an aliphatic

head and an aiiphatic tail. The aliphatic head, a complex heterocycle, is believed to be the

crucial part of the biotin rnolecule responsible for binding to airidin proteins (Green, 1975). No

physiological compound other than biotin binds to avidins with any significant strength

(Green, 1975). Biotin is an endogenous molecule, also found in animal and human tissues. In

humans, the concentration of circulating biotin is estimated to be 0.5 nglmL (Mock and

Dubois, 1986) and in mice. is approximately 8-fold greater at 4 ngIrnL (Rosebrough et al.,

199 1).

There is no evidence so far indicating that avidins may be tonic to humans, horvever,

they do appear to be imrnunogenic, which is not an unespected finding given their xenogencic

nature. Hnatowich et al. ( 1985a) reported that 81 10 patients administered streptavidinylated

mAb displayed a human anti-streptavidin response (HASA), as well as a HAMA responsc.

Paganelli et al. (199lb) also reported the development of HASA responses appearing 15-30

daps following streptavidin administration in 7/20 patients in the study.

Biotinylated mAb and Radiolabeled (Strept)avidin (2-step)

Espenmental evidence indicates that biotinylation of mAbs does not affect their

biodistri bution o r elimination [rom blood in animals and patients (Sini tsyn e t al., 1989;

Paganelli et al., 1 WOa; Saga et al., 1994). The rate of tumour uptake and intratumoural

distribution of biotinylated mAb should be similar to that of native mAb since biotinylation

does not increase the size of the IgG molecule significantly. Saga et al. ( 1994) found the

distribution of biotinylated D3 mAb in lung metastases to be similar to that of native D3 mA b.

Biotinylation of mAbs can circumvent the loss of immunoreactivity d u e to radiolysis, a

problern which has been reported for certain radiolabeled mAbs (Nikula e t al., 1995).

However. radiolysis of radiolabeled streptavidin is a possibility, that could adversely affect its

binding to biotinylated mAb. Similarly. ndiolysis of radiolabeled biotin could also occur and

affect its binding to streptavidinylated mAb. Paganelli et al. ( 1990a) reported no loss in

imrnunoreactivity as a result of the biotinylation of AUA 1 mAb. Mathematicai modeling has

projected that plasma clearance of streptavidin would be approximately 3-fold faster than that of

intact mAb (Sung et al.. 1994). Paganelli et al. ( lW) reported that streptavidin. labeled wi th

indium 1 I l ( W n ) , had a fl-phase half-life of 45 hours in ovarian cancer patients, following ip

administration. However. when administered by i v injection. 1 1 'In-labeled streptavidin

displayed a half-life in the blood of 16 hours (Paganelli et al., 199 la). Animal studies

demonstrated that the biodistribution and phmacoknetics of streptrividin differ substantially

from that of avidin (Schechter et al.. 1990; Rosebrough, 1993), which could have important

implications in dinical pre-targeted RIT using biotinylated mAb and radiolabeled (strept)avidin.

The elimination of streptavidin from blood is markedly slower than that of avidin. and unhkc

avidin, accumulates in the kidneys. Moreover. different preparations of streptavidin display

markedl y different biodistri bution. at least in mice (Schech ter et al., 1990). These authcrs

observed that the truncated form of streptavidin, which is present in most commercial

preparations, had an overail retention in blood and normal tissues. escept kidney. that \vas

approsimately 3-lold lower than that of intact streptavidin. However. truncated streptavidin

eshibited approsimately an 8-fold increase in hdney localization cornpared CO intact forms of

the protein.

Tumour upiake of radioactivit) using biotinylated mA b and radiolabeled streptavidin

can be superior than that achieved using radiolabeled mAb. In a study by Paganelli et al.

(1990a). tumour radiolocalization at 24 hours following ip administration of radiolabeled

streptavidin to tumour-bearing mice. pre-targeted with biotinylated mAb, was Cfold greater

using the 2-step approach cornpared to that of radiolabeled mAb. In addition, blood levcls of

radioactivi ty were approximatel y 3-fold lower for radiolabeled s treptavidin compared to

radiolabeled antibody. However, increased liver uptake with the 7-step approach was observed

due to the formation of immune complexes between circulating biotinylated mA b and

radiolabeled streptavidin. In an animal model, Saga e t al. ( 1994) reported that the rate of

tumour uptake of 1251-streptavidin in lung metastases, pre-targeted with biotinylated D3 mAb.

was faster than that with conventional targeting with i 3 V D 3 mAb. However, i t also appeared

that the "binding s i te bamer" phenomenon may have prevented deep penetration of the

streptavidin molecule in the turnour nodule. Given the very high binding affinity between

streptavidin and biotin (Ka= 1015 M-1). intratumoural distribution of radiolabeled streptavidin

rnay be more heterogeneous than that of the biotinylated mA b, wi th its binding limi ted to the

biotinylated mAb adjacent to tumour vasculature. Mathematical modeling, in fact, predicts that

the degree of intratumoural heterogeneity arising [rom the 2-step approach could be p a t e r

under some conditions (van OsdoI e t al., 1993).

(S t rep t )av id inyla ted mAb and Radiolabeled Biotin (2-step)

In this al ternative 2-s tep approach, the radiolabeled effector molecule, biotin, is more

npidly eliminated from the blood than streptavidin. Paganelli et al. (1990b) determined, in

patients, the hall-lire of biotin in blood to be approsimatel y 3 minutes, ivhen administered by iv

injection. In the first 24 hours, up to 80% of the injected dose was escreted in the urine, while

kidney uptake ivas only 3 8 of the injected dose. Similarly, Kaloionos et al. ( 1990),

determined the a-phase half-life of biotin, in serum. to be 7.4 minutes and the p-phase half-life

to be 4.7 hours. Within 34 h o u n after administration, 71% of the injected activity was cscreted

in the urine. An advantage of using strepiavidinylated mAbs and radiolabeled biotin is the

potential amplification in the delivery of radioactivity to the tumour because of the terravalent

binding between (strept)avidin and biotin. Furthemore, unlike (strept)avidin, biotin does not

appear to sequester in any normal tissues, at least when conjugated to DTPA (Rosebrough,

1993, Paganelli e t al., 199 1 b). However, preliminary data suggest that the biodistri bution of

radiolabeled biotin depends o n the physico-chernical properties of the chelating agent, ie.

charge and lipophilicity (Paganelli et al.. 1991a).

The presence of endogenous biotin may be a potential complication of using this 3-step

approach since endogenous biotin may saturate the biotin binding si tes of streptavidi nylated

mAb in the tumour. Rusckowski et al. ( 1997) demonstrated that endogenous biotin in mice can

significantly interfere with binding of radiolabeled biotin to the tumour. in tumour-bearing mice

pre-treated with streptavidin to reduce endogenous biotin, streptavidinylated mAb localized in

tumour senografts was capable of binding 95% of the adrninistered radiolabeled biotin. In

contrat. without pre-treatment, only 6 4 of radiolabeled biotin bound to streptavidinylated

mA b. The presence of endogenous biotin, hoivever, ma? not be as significant in humans since

the concentration of circulating biotin is &fold less compared to that in mice (Rosebrough et

al., 1991; Mock and Dubois. 1986). In a clinical study by Kalofonos e t al. (1990).

endogenous biotin did not saturate the biotin binding capacity of streptavidinylated mAb in

75% of the patients. Similm results were aiso found by Paganelli et al. ( 1991 b). Administering

higher doses of radiolabeled biotin may be an alternative strategy to circumvent any potential

complications of endogenous biotin.

Mathematical modeling(van Osdol et al.. 1993; Sung and van Osdol, 1995) predicted

that, at similar doses, the intraturnoural distribution of strepcavidinyiated mAb ivould be more

heterogeneous than that of native mAb and biotinylated mAb because of i t s higher molecular

weight. The addition of one streptavidin rnolecule to the mAb would increase its molecular

weight from 150 kDa to approsimately ? 15 Da, whereas biotinylation does not sipficanrl y

increase the molecular weighr of the IgG rnolecule. Sung and \,an Osdol ( 1995) also predicted

that the rate of tumour uptake of radiolabeled biotin would be considerabl y faster than that of

native mAb and streptavidin because of its much smaller molecular weight. However, biotin

ma? also be subject to the effects of the "binding site barrier" in the presence of pre-targeting

wi th streptavidi nylated mAb (van Osdol et al., 19%). Several invcstigators have reported no

significantiy large losses in rnAb immunoreactivity as a result of streptavidinylation (Kalofonos

et ai., 1990; Ngai et al., 1995).

Axworthy et al. (1995) treated tumour-bearing mice with streptavidinylated mAb

followed by, a clearing agent. and goY-DOTA-biotin, at doses ranging from 200-800 pCi.

More than 90% of the administered radioactivity was eliminated frorn the body within 24 hours

following injection and no apparent myeloto'ricity was obsewed. Complete turnour regression

lasting for 2200 days in 80% of mice bearing subcutaneous hurnan breast turnours and in

1 ûûQ of mice bearing subcutaneous human colorectal turnours \vas obsen-ed. The absorbed

radiation doses delivered to the colorectai turnours were estirnated to be > 150 Gy1800 pCi. 1 n

contmt. conventional RIT with gV-mA b demonstrated greater tosicity and lower thenpeutic

efficacy atequivalent doses up to 400 pCi. At 400 pCi, however. death in 1 0 % of the mice

was observed.

Biotinylated mAb, (Strept)avidin, and Radiolabeled Biotin (3-step)

The optimum time betaeen administration of the bifunctional anti body and effector

molecule will depend on how fast the anti body is eliminated from the blood and normal tissues.

In the 3-step approach, elimination of biotinylated rnAb from the blood can be encouragd by

adrninistering unlabeled (strept)avidin as a clearhg agent. The (strept)avidin acts as a h a s e

molecule. binding to circulating biotinylated mAb and forming immune cornpleses that are

npidly sequestered by the liver. In addition, (strept)avidin binds to biotinylated mA b localized

in the turnour, thereby providing binding sites for the radiolabeled biotin to be administcred in

the third step. The 3-step approach has k e n successful in tumour imaging studies. In turnour-

bearing mice, the 3-step approach achieved a T: BLD ratio of 50: 1 at only 7 hours d e r injection

of radiolabeled biotin, while that achieved by biotinylated mA b and radiolabeled streptavidin at

4 hours follow ing injection of radiolabeled streptavidi n was approsimatel y 1 2- fold loiver.

nnging from 3.5: 1 to 4: 1 (Paganelli et al.. 1990a). Moreorer. the tumour:liver ratio was

between 30: 1 to 40: 1 [or the 3-step approach, whereas that of the 7-step approach was 4: 1.

Unlike with biotinylated mAb and radiolabeled streptavidin, very little liver uptake of

radioactivity was observed in the 3-step approach perhaps because the sequestered immune

complexes were metabolized within the hepatocytes and unavaiiable for binding to radiolabeled

biotin.

Cremonesi et ai. (1997) used the 3-step approach to treat forty-five glioma patients.

Patients were administered biotin ylated mA b folloured by 3 clearing agents, avidin,

streptavidin, and biotinylated human semm albumin. In the third step, patients were

administered gV-DOTA-biotin at a dose of 25-75 mCi/m2 mised with tracer amounts of 1 1 f n-

DOTA-biotin to moni tor the biodistn bution of DOTA-biotin. Elimination of DOTA-biotin was

rapid with 65% of administered radioactivity recovered in the urine uithin 34 hours. At 3

months pst- therapy 56% of the patients demonstrated stable disease, 24% demonstrated

progressive disease, and 30% had a complete response. Houever, a t 13 months pst-therapy,

84% of the patients demonstrated progressive disease and the percentage of patients

demonstrating a complete response decreased to 16%. T h e absorbed radiation dose to the

turnour ulas estimated to be 0.15 + 0.081, GylrnCi, while the dose to the bone marrou was

0.009 + 0.005 GytmCi. Al though a majori ty of the patients displayed hematological toxicity,

the myelosupression observed was mild. Compared to the average radiation dose delivered to

the tumour, the average doses delivered to other normal tissues were relativeiy low at 0.038 c

0.015, 0.015 2 0.013. and 0.006 + 0.003 GylrnCi for Iiidneys, liver. and brain, respectively.

The 3-step approach possesses the a d w m g e s offered by both 2-step approaches. 1 t

utilizes the srnaller sized biotinylated mAb, which theoretical models have predicted to have

superior tumour distri bution compared to streptavidinylared mA b. Secondl y. the 3-step

approach also utilizes biotin as the effector molecule, which has demonstrated to have a shoner

half-life in blood, less cross reactivity with normal tissues, and a faster rate of tumour uptake,

cornpared to streptavidin. Hoivever, the 3-step approach is more cornples than 2-step prc-

targeting because of the two, rather than one, tirne intemals to consider, as well as an additional

reagent for ahich an appropriate dose must be determined.

Pre-targeting with Streptavidin-CC49 Monoclonal Antibody and DOTA-biotin

CC49 Monoclonof Anribody

A first generation mAb directed against TAG-73 antigen, B72.3, was generated by

Colcher et al. ( 198 1) by using a membrane-ennched fraction of a human breast carcinoma

biopsy as the immunogen. The TAG-72 antigen is a high rnolecular weight mucin-like antigen

expressed in >94% of human colorectal cancers and a majority of ovarian, breast. and non-

small ce11 lung carcinomas (Johnson et al., 1986; Thor et al.. 1986). Muraro et ai. (1988)

purilied TAG-73 from the LS174T human colon cancer ce11 line and used the antigen to

generate a series of twenty-eight second generation IgG mAbs. designated "CC" for colon

cancer. The second generation mAb. CC49, demonstrated superior reactivity to TAG-72 as

compared to 872.3. The Ka of CC49 is 1.6 n IO1 M- l . whereas that of B72.3 is nearly 6-fold

lower at 2.5 s IO9 M-1. In the animal model. CC49 rnAb has displayed superior turnour uptake

and anti-iumour effects compared to B72.3, possibly as a resul t of i ts higher Ka (Schlom et al..

1992; Colcher et al.. 1988). In a phase I RIT study, Divgi et al. (1995) treated twenty-four

colorectal cancer patients with l 3 lI-CC49 mAb at a dose of 15-90 rnCi/m2. Ninetÿ Cive percent

of known lesions were visualized by I3l1-CC49 mAb, thereby qualitatively demonstrating

escellent tumour targeting by the ndioantibody. Despite tumour localization, however, no

major responses were observed. Nevertheless, sis patients demonstraied stable disease for 4

weeks p s t - therapy and subsequently received a second dose of I 3 1-CC49 rnA b. As a resul t.

one patient had a minor response following the second treatment. Human anti-mouse antibody

responses and dose-related hematological tosicity were observed. In a phase I I RIT tnal

(Murray et al., 1994), fifteen patients were treated with 75 mCi/m2 of 1311-CC49 mAb.

Tumours were successfully imriged in all patients, however, no objective responses were seen.

Absorbed radiation doses to the tumours were estimated to be between 0.19-66.7 Gy and

HAMA and hernatological touicity resul ted from treatment

Sîreptavidinylnri011 of CC49 Manoclonal Aiitibodv

Streptavidin has k e n conjugated successfuily to CC49 rnAb in our lab (Ngai et al..

1995; Alvarez-Diez et al., 1996). Ngai et al. ( 1995) evaluated three different approaches to

conjugate streptavidin to CC49 mA b. One approach involved conjugating the CC49 mA b wi th

sulfosuccinimidyl4(N-maleimideomethyl)cyclohesane- 1 -carbox y late (su1 fo-SMCC) via

lysine residues on the mAb. The maleimide-derivatized rnAb was then reacted with chemically-

thiolated streptavidin to generate the streptavidin-CC49 rnAb conjugate. For the second

conj ugation approach, CC49 mA b was biotinylated and then reacted wi th streptavidin.

Biotinylation was performed by reacting biotin LC-hydrazide with osidized carbohydnte

moieties on the Fc region OC the mAb. In the third method, disuiphide moieties on the CC49

mA b were reduced and then reacted wi th malei rnide-derivatized streptavidin to produce the

conjugate. Strepfavidinylation in al1 three approaches did not large1 y affect the

immunoreactivity of the CC49 mAb. The third approach provided the best conjugate yield at

95%. ie. 4% free streptavidin, and produced a conjugate with a molecuiar weight of

approsimatel y 3 10 ma, as determi ned by size-escl usion hi@ pressure l iquid chromatograph y

(HPLC). This sugpsted that conjugation of only one streptavidin molecule (65 D a ) per CC49

mAb ( 1% kDa) was achieved. This conjugation procedure was subsequently used by Alvarez-

Dicz et ai. (1996) in a ?-step pre-targeted radioimmunoimaging study, in which the.

successfully imaged subcutaneous LS 174T human colorectal senografts grown in athÿmic

mice.

DOTA- biotirt

1, 4, 7, 10- tetraazacyclododecane-N,N'.N",NM'- te traacetic acid (DOTA) is a syn thetic

macrocyciic chelatorsuitable for chelating ndiornetals such as, g o y , i l l ln , and 67C~, and has

k e n reported to fom radiometal-chelator complexes more stable than those formed with other

commoni y used chelators such as, DTPA (Moi et al., 1990). Over an 18 day period in vitro, no

measurable loss of yttrium 88 (*8y) from DOTA was observed in semm under physiologd

conditions. In contrast, a loss of OS7%lday was observed when 88Y was chelated with I-p-

nitrobenzyl-DTPA. An even more dramatic loss of lO%lday was obsewed with the

monobutylarnide-DTPA derivative. Deshpande et al. ( 1990) also observed no rneasurable loss

of g q in vitro from DOTA-conj ugated Lym- 1 mA b in serum under physiolo@cai conditions.

The NeoRv Corporation of Seattle, Washington has developed a DOTA-biotin molecule

with a rnolecuiar weight of 900 Da for use in pre-targeting strategies (Su et al., 1995) (Figure

4). The DOTA-biotin compound labeled wi th has demonstrated rapid blood clearance in

tumour-bearing mice pre-targeted with streptavidinylated mAb (Axworthy et al., 1995) (see:

"Streptavidinylated mAb and Radiolabeled Biotin (2-step)") and in a clinical pre-targeted RIT

study by Cremonesi et al. (1997) (see: "Biotinylated mAb. (Strept)avidin, and Radiolabeled

Biotin (3-step)"). This same compound is also currently king used by our laboratory in

cvaluating streptavidin-CC49 and 90Y-DOTA-biotin for pre-targeted RIT of LS 174T human

colorectai uenografts grown subcutaneousl y in athymic mice ( Domingo and Rei 11 y, 1997).

RESEARCH OBJECTIVES

Statement of the Problern

Bone mmow toxicity is a major limitation of RIT because of the relativelp long half-life

of the radiolabeled antibody in blood. Moreover, bone marrow toxicity is dose-limiting,

thereby making it difficult to deliver effective radiation doses to the tumour. A reduction in

bone rnarrow toxicity can potentially enable higher doses of radioactivity to be sarely

administered to the patient and possibly, improve the radiation dose delivered to the tumour

tissues.

Objectives of Researc h

T h e objective of my reswch is to evaluate streptavidin-CC49 and "Y -DOïA-biotin 10r

pre-targeted RIT of subcutaneous LS l74T human colorectal senografts grown in athymic

mice. This will be accomplished by: i) evaluating the biodistribution of DOTA-biotin in

BALBlc mice (Esperiment 1); ii) establishing a dosing regimen for streptavidin-CC49 and

DOTA- biotin for pre-targeted RIT of athymic mice bearing LS 174T colorectal senografü:

(Esperirnent II); and iii) evaiuating the therapeutic efficacy and tosicity of streptavîdin-CC49

and gm(-DOTA-biotin administered to athymic mice beanng LS 174T colorectal senoprafts

( Esperiment III).

EXPERllMENT 1:

BIODISTRIBUTXON OF DOTA-BIOTIN

Introduction

Our laboratory has successfull y imaged su bcu taneous human LS 1 74T colorectal

xcnografts, grown in athymic mice, using a pre-targeting approach with streptavidin-CC49 and

lh-DTPA-bi-tin (Alvarez-Dicz et al., 19%). Rather than using DTPA-biocytin for our

evaiuation of pre- targeted RIT wi th streptavidin-CC49, we have chosen to utilize a DOTA-

biotin compound developed by the NeoRx Corporation (Su et al ., 1995). Zn vitro and in vivo

studies have demonstrated that DOTA chelators can f o m a more stable comples with

radioyttrium than can chelators such as, DTPA (Moi et al., 1990; Deshpande et al. 1990;

DeNardo et al., 1994). Dissociation of 90Y from its chelator in vivo and subsequent

accumulation of the free radionuclide in cortical bone is an important concern in RIT, because

of the reported correlation between myelotosicity and bone uptake of free 90Y ions (Stewart et

al., 1988; 1990; Hnatowich et al., 1985b). Stewart et al. ( 1988) proposed that bone uptrùce of

goy, can in Tact, be a dominant source of bone rnarrow irradiation and tosicity. In a pre-clinicd

comparative toxicity study, a 210 pCi dose (LDSO) of 90Y -DTPA-BrE-3 mA b delivered a

higher radiation dose to the bone marrow than did 308 pCi (LDSo) of 'OY-DmA-BrE-3 rnAb

(DeNardo et al., 1994). These differences were attributed to a higher bone uptake of free 9flY,

rvhcn adrninistered in the f o m of 90Y-DTPA-BrE-3 mAb. Hence. in light of espenmental and

clinical evidence, DOTA-biotin was chosen as the effector molecule for our evaluation of pre-

targeted RIT, in order to minirnize the deposition of free 90Y in bone and an)- bone marrow

tosici ty that may arise as a result.

However, apart from forming stable radiornetal-chelator complexes, the DOTA-biotin

molecule should also display the foilowing behaviour iri vivo : i) npid elimination from blood,

primarily through rend escretion and ii) little or no affinity for normal tissues, in order to be a

suitable effector molecule for pre-targeting stntegies. The objective of this experiment,

therefore, was to evaluate the biodistribution of the DOTA-biotin chelator in non-pre-targeied

mice and vcrify that it displays those properties described above. Because of the similar

chemisty betwcen 1 1 1In and 90Y and the difficul ty in obtaining accunte measurements of

radioactivity in tissues using gamma or beta-scintillation counten (Roselli et ai., 1989). others

have used 1 l'In as a tracer for 90Y in biodistribution studies (Leichner et al., LW; Parker et

al., 1990; Cremonesi e t ai.. 1997). T o moni tor the biodistri bution of DOïA-biotin. we labeled

it with 'In with the reaiization that, despite the similar chemistry between 1 1 lin and goy. the

stability of the 1 1 1111-DOTA comples may be slightly different than that of the 90Y-DOTA

cornplex. Consequently, the biodistn bution of DOTA-biotin. determined by the 1 In

radioactivity measured in tissues, rnay be slightly different than that of DOTA-biotin when

labeled with 90Y, due to differences in biodistribution between the two radionuclides (Durbin et

ai., 1960; Roselli et al., 1989).

Methods

Prepnratiori of 1 h-DOTA-biotitz ( s e : A ppendis 1 V)

DOTA-biotin was radiolabeled with IlIn chloride to a specific activity of 50 vCi /~g .

Indium 1 1 1-DOTA-biotin was diluted for injection with stenle 0.9% saline. The radiochemical

puri ty of 1 1 In-DOTA-biotin assessed by reversed-phase chromatograph- was 974.

Bionistriblitiori of DOTA- biuriri

Fifteen four-week old female BALBlc mice (Charles River Laboratories. Boston. MA)

were acclimatized in The Toronto General Hospital Animal Facility for two weeks upon amval.

Each rnouse received 7 pg ( 100 pCi) of 'In-DOTA-biotin by tail vein injection following

sedation wi th subcutaneous injections ( 100 pL) of an anesthetic cocktail containing,

2S%:(wlv) ketamine (RogarISTB Inc.. London. Ont.), O. 1 I%:(wlv) sylazine (Miles Canada

1 nc.. Etobicoke, Ont.), and O.O?-5%: (wlv) aceprornazine ( Ayerst Laboratories, Montreal,

Québec). Groups of 2-3 mice were sacrificed at 1,?, 3 , 6 , and 12 houn following injection of

1 In-DaA-biotin and their tissues (blood. hdneys. liver, stomach. intestines, spleen, liver,

heart, lungs, and right femur) collected. Tissue sarnples were weighed and counted for 1 'In

ndioactivi ty in a gamma well counter (Packard Mode1 Auto-Gamma 56M, Downer's Grove,

IL) using a window of 150260 keV to include the 177 keV and 247 keV y-photons of l In. A

sample of the l 'In-DOTA-biotin injectate was aiso counted with the tissues to correct for

physical decay. The biodistribution of DOTA-biotin was espressed as the Q id/g of tissue,

Results

One mouse died from anesthetic overdose prior to injection with 1 l 'In-DOTA-biotin,

and therefore, was removed from the study. Indium 11 1 radioactivity was rapidly eliminated

from the b l d a i t h 4% idlg remaininp at 1 hour pst-injection (pi) of 1 l lIn-DOT'A-biotin. At

2 and 12 hours pi, levels of ndioactivity decreased by 5- and 100-fold, respectively (Figure 5).

There was very little upialie o l radioactivity in normal tissues such as. hem, lung, liver,

spleen, and femur (Figure 6). Moreover, the radioactivity in these tissues aas rapidly

eliminated dunng the first 12 hours foilowing injection of 1 1 1 In-DOTA-biotin. From 1 hour pi

to 12 hours pi, levels of radioactivi ty decreased by 40-, 7Q. 60-. and 70-lold in the hem.

lungs, liver. and femur, respectively. Kidneys had the highest levels of 1 I l In radioactivity at

3.2% id/g at 1 hour pi, decreasing 23-îold to 0.4% idlg at 12 hours pi. which aas still higher

than other normal tissues, escept intestines. In the intestines, there appeared to be some

retention of 1 1 1 In radioactivity n i th levels decreasing only slightly from 0.5% id/g at 1 hour pi

to 0.4% id/g at 12 hours pi. The 9 id of l l I In radioactivity per whole tissue was also

determined and is presented in Table 4.

Discussion

We have chosen to use a DOTA-biotin compound deïeloped by the NeoRx Corporation

(Su et al., 1995) for our evaluation of pre-targeted RIT with streptavidin-CC49. DOTA

chelaton have been shown in vitro and itz vivo to fom a more stable comples with yttrium than

chelators such as. DTPA (Moi et al.. 1990; Deshpande et al.. 1990; DeNardo et al., 1994).

Time pi (hours)

Figure 5 . Elimination of DOTA-biotin frorn blood. DOTA-biotin (2 pg) was iabeled with "'In (100 pCi) and administered by tail vein injection to BALBk rnice. The mice were sacrificed at various time points follouing injection. Data points represent the mean % id/g of whole blood + SD (n=3 except at 12 hours. n=2). Inset: data ploned on semi-log scale and fined to dual phase exponentiai decay equation (GraphPad Prism sofhare Version 2.0, GraphPad Software Inc.. San Diego. CA 1.

u . - / -

Abbreviations: pi. post-injection; % idg % of injected dose per gram.

Tissue

Figure 6. Tissue distribution of DOTA-biotin. DOTA-biotin (2 pg) was labeled with "'In (100 pCi) and administered by tail vein injection to BALB/c mice. The mice were sacrificed at various time points following injection. Bars represent the mean % id/g of tissue f SD (n=3). " n=2. Abbreviations: pi, post-injection; % idlg, % of injected dose per gram; BLD, blood; KID. kidneys; HRT, heart; L W . lungs; LW. liver; SPL, spleen; STO, stomach; MT. intestines; FEM, femur.

Minimizing the dissociation of 90Y in vivo from its chelator is important in keeping the

radiation dose delivered to the bone marrow and the risk of myelotoxicity as low as possible

(DeNardo et ai., 1994). In addition. the effector molecule should be rapidly eliminated h m

blooci, pnmarily by renal eacretion, and should demonstrate very little uptake in normal

tissues. We evaluated the biodistnbution of DOTA-biotin, labeled with 'In, in BALBk rnice

at 1,3,3,6, and 12 hours pi to verify that it exhibits those propenies described above.

According to the biodistribution of 1 1 'In radioactivi ty, the DOTA-biotin chelator was

npidly eliminated from the b l d (Figure 5) and had very little uptake in normal tissues (Figure

6). Chinol et al. (1997) evaluated the biodistribution of DOTA-biotin, labeled with 90Y, in

turnour-bearing mice pre-targeted with biotinylated FO'023C5 mAb and avidin. The DOTA-

biotin compound used by Chnol et ai., however, appears to have a chemicai structure diffcrent

to that of the DOTA-biotin used in this study. The molecular weights of both DOTA-biotin

cornpounds appear to be similar but, the DOTA-biotin cornpound used by Chinol ct al. is

apparent1 y susceptible to cleavage between the DOTA and biotin moieties, whereas the DOTA-

biotin used in this study is not (Su et al., 1995). Nevertheless, results from b t h studies are

comparable in that both DOTA-biotin compounds demonstrated relatively very little uptake in

normal tissues. In Our study, uptalie of 1 1 1In radioactivity in normal tissues, including the

fëmur, waç <0.3% idlg at 3 hours pi, and in Chinol e t al., normal tissue uptake of

radioactivity was approximately 4 . 2 % idlg at 4 hours pi. Chinol et al. (1997) reported ihat

90% of the injected activity was escreted into the urine at 4 hours pi oc 9OY-DOTA-biotin.

Although we did not measure for the presence of l l l1n excreted into the urine, the pattern ol'

uptake and elimination of 1 1 'In in the kidneys suggest that DOTA-biotin was climinated by

rend excretion. In this study, there was some retention of radioactivity in the intestines,

suggesting that the DOTA-biotin compound rnay have also been eliminated through

hepatobiliary excretion. On the other hand, the retention of radioactivity rnay have represented

elimination of free l l h , released from DOTA-biotin, through the gastrointestinal tract

(Durbin, 1%0).

This is not the first study to evaluate the biodistribution of the DOTA-biotin compound

developed by the NeoRx Corporation. In a pre-clinical RIT study conducted by NeoRx

(Axworthy e t al.. 1995)- tumour-bearing mice were pre-targeted with streptavidin-mA b,

administered a clearing agent, and then injected with -DOTA-biotin. More than 908 of the

administered 90Y-DOTA-biotin was eliminated from the body within 24 hours pi. Moreover,

no myelotosicity was observed following treatment with 200-800 pCi of goY-DmA-biotin. In

a ciinical study, Cremonesi et ai. ( 1997) treated forty-five gliorna patients wi th a 3-step pre-

targeting protocol. Patients were administered biotinylated mAb followed by, avidin and

streptavidin, and -DOTA-biotin. To moni tor the biodistri bution of DOTA-biotin, the final

preparation was spiked with DOTA-biotin labeled with 1 11 n. Approximatei y 65% of' the

adrninistered 'In radioactivity was excreted into the urine and the radiation doses delivered to

normal tissues from g0Y were low.

T h e results from our study must be interpreted with the realization that the

biodistribution of DOTA-biotin. labeled with l 'In, may be slightly different than that of

DOTA-biotin labeled wi th g o y . Roselli et al. ( 1989). however. demonstrated that substi tuting

1 1 'In as a tracer for 90Y can accurately predict the biodistribution of an 9OY -1abeled species, if a

suiiable chelator is used. Using a chelator that forrns radiometai-chelator corn pleses wi th both

1 1 1 In and at equal stability, may lead to similar biodistribution when either radionuclide is

used. The stabili ty of 1 In-DOTA and 90Y -DOTA cornpleses was not evaluated for this study.

Nevertheless, this study demonstntcd that DOTA-biotin is rapidty eliminated from b h d ,

undergoes rend escretion. and has very little uptake in normal tissues, thereby indicating that,

in most respects, i t is suitable for pre-targeting strategies. Resul ts from previous pre-clinical

(Chinol e t al., 1997; Asworthy et al., 1995) and c l i n i d (Cremoncsi et al.. 1997) studies

further support Our conclusions.

EXPERIMENT XI:

DE~LOPMENT OF A DOSING SCHEDULE FOR PRE-TARGETED RADlOIlMlMUNOTHERAPY

WITH STREPTAVIDIN-CC49 MONOCLONAL ANTIBODY AND goy-DOTA-BIOTIN

Introduction

Successful pre-targeted RIT requires that certain parameters be optimized, particularly

those constituting the dosing schedule to be employed. These include: i ) the dose of mA b

conjugate to be administered; ii) the tirne interval(s) between the administration of the mAb

conjugate, clearing agent (if applicable), and effector molecule; and i i i ) the dose of effector

molecule to be administered, ie. mass and radioactivi ty. Ideall y, in a 3-step approach, enough

mAb conjugate should be administered so that maximum tumour uptake of the antibody is

achieved. Second1 y, the time interval between the administration of the mA b conjugate and

effector molecule should be one that allows for sufficient elimination of the antibody from the

blood and normal tissues to occur, and in addition, one that maintains as high a level of

antibody in the tumour as possible. Lastly, the administered dose of the effector rnolecule

should saturate al1 of the available hapten binding sites in the tumour in order to ma~imize the

radiation dose delivered to the tumour. Previously in our labontory, Alvarez-Diez ( 1995)

determined that a 2% ~g dose of streptavidin-CC49 achieved higher absolute tumour uptake ot'

antibody than did smaller doses of 50 and 100 pg. No dose higher than 250 ~ ( g was evaiuated

but, i t was possible that greater tumour upiake of antibody may have been achieved at higher

doses. Nevertheless, i t was decided to continue using the 250 [cg dose in Our evaluation OC pre-

targeted RIT.

Determining the optimum parameters for pre-targeted RIT must be donc experimentally,

and therelore, the aim of this study was to develop a dosing schedule that will be subsequentiy

employed in a pre-targeted EUT approach to treat athymic mice, bearing subcutaneous hurnan

LS 1 74T colorectal senografts, wi th streptavidi n-CC49 and 9OY -DOTA- biotin. The speci fic

objectives were to determine the time interval and dose of effector molecule that rneet the

criteria descnbed in the preceding paragraph. To determine these parameters, ive performed

dual label biodistribution studies in LS 174T tumour-bearing mice, labeling streptavidin-CC49

with 1251 and DOTA-biotin with 1 ' 1 In. Two different time intervals (40 and 72 hours) and four

different doses of DOTA-biotin ( 1,20,40, and 150 pg) were evaiuated.

Methods

Radiola beling of Streptavidin wiîh 2 251

Iodine 125-streptavidin was prepared to a specific activity of 1 yCi/pg according to the

Iodogen method (Fraker and Speck, 1978). Streptavidin (10 mg/mL) (Sigma Chemicai Co.,

St. Louis, MO) was incubated with Na12jI (Amersham. Oakville. Ont.) for 10 minutes at room

temperature in a glass reaction v id coated with 30 pg of 1, 3, 4, 6,-tetrachloro-3a. 6a.

diphenylglicounle (Iodogen) (Sigma Chernical Co.). At the completion of the radiolabeling

procedure. 1 251-streptavidin was purified from free l29 on î. Sephades G-50 mini -column

eluted with 0.9% saline (see: Appendix II for preparation of Sephades G-50 mini-columns).

The radiochernical puri ty of 1 251-streptavidin was detemined by paper chrornatcgnphy usi ng

Whatman #1 chromatography paper. Radiochernical purity of 125Lstreptavidin was 2974.

Preparntio~z of (1251-Sîreptavidin)-CC49 Motzocfonaf Antibodv

CC49 mAb was purified from mouse ascites fluid and analyzed for purity by sizc-

esclusion HPLC and sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-

PAGE) (see: Appendix 1). CC49 mA b was then conjugated to 1 25Lstreptavidin and anal yzed

for purity and molecular weight by size-exclusion HPLC and SDS-PAGE ( sec Appendis II).

The immunoreactive fraction of (1251-streptavidin)-CC49 preparations ranged [rom 50-904

(sa: Appendis III).

Preparuliorz a/' 1 1 Irz- DOTA- bio fit1 (see: A p pendix 1 V)

DOTA-biotin was labeled with l l1n chloride. The radiochernical puri ty or 1 1 11n-

DOTA-biotin assessed by reversed-phase chromatognphy was 294% Ninety one to 94% of

1 1 1 In-DOTA-biotin bound to streptavidin in viiro. Indium 1 1 1 -DOTA-biotin was diluted for

injection with sterile 0.9% saline.

Biodistrîbution of S~eptavidin-CC49 Monoclonal An fibody and DOTA-biotin

LS 174T human colon cancer celis were injected subcuianeously in eleven six- to eight-

week old female athymic mice and the turnour xenografts allowed to grow for 8- 15 days, at

w hich tinte tumour diameters ranged from 3-6 mm (see: Appendix V). In order to determine an

appropriate time intenral between the administration of streptavidin-CC49 and DOTA-biotin for

the pre-targeted RIT study, the LS 174T turnour-bearing mice were administered 250 pg (50-60

pCi) or (I~5i-streptavidin)-CC49 by ip injection followed by 1 pg ( 100 pCi) of lin-DOTA-

biotin, administered by tail vein injection, at either 40 or 72 hours pi (n=5-6 miceltime interval)

of ('25Lstreptavidin)-CC49. Prior to receiving tail vein injections of 1 1 l In-DOTA-biotin, the

mice were anesthetized as previousl y descn bed in Experiment 1. Sis hours later, the mice were

sacrificed and their tissues (tumour, blood, hdneys, liver, stomach, intestines, spleen, liver.

heart, lungs, and right femur) collected. Two additional groups of LS 174T tumour-bearing

mice (n=6 micefgroup) served as controls and received either 1251-streptavidin alone (50 yg;

M-60 1tCi) by ip injection or l 'In-DOTA-biotin aione ( 1 pg; 100 pCi) by tail vein injection.

Mice receiving 1251-streptavidin and 1 l 'In-DOTA-biotin were sacrificed at 46 and 6 hours pi of

ndioactivi ty. respective1 y, and their tissues collected. To determine the appropriate dose of

DOTA-biotin required for the pre-targeted RIT study. eighteen LS 174T tumour-bearing mice

alerc administered 150 pg (50-60 pCi) of (*251-streptavidin)-CC49 by ip injection. Rcsults

from the time intemal studp indicated that 40 houn was the appropriate time interval (sec:

Results and Discussion). Therefore. the mice received ei ther 10, 40, or 1 M yg ( 100-300 pCi)

ol' 1 1 Iln-DOTA-biotin by tail vein injection (n=6 mice/dose) 40 hours pi of (1251-strept.avidin)-

CC49. The mice were sacrificed at 6 hours pi of l l h-DOTA-biotin and their tissues collected.

Al1 tissues were weighed and counted for lin and 1 2 9 activity in a gamma well

counter (Packard Model Auto-Gamma 5650) using windows of 150-260 keV and 15-80 keV,

respective1 y. Al1 measuremen ts of 1z51 radioactivi ty were corrected for 1 1 1 In spillover.

Samples of the ( l 251-streptavidin)-CC49 and l l In-DOTA-biotin injectates were ai so counted

with the tissues to correct for physical decay. The biodistnbution of streptavidin-CC49 and

DOTA-biotin was espressed as the % id/g of tissue, calculated accordingly as previously

described in Experiment 1.

Statisticnl Anaiysis

Mean tissue uptake of radioactivi ty was compared between groups using the Student's

unpaired t-test wi th a level of significance of @.O5 Mean DOTA-bi0tin:streptavidin-CC49

ratios were compared between groups using one-way analysis of variance with a level of

significance of p4.05.

Resu its

Size-exclusion HPLC of CC49 mA b purified from mouse ascites fluid showed a single

pcak with retention times of approsimately 8.0-8.5 minutes, which by extrapolation from a

standard c u n e of molecular weight standards and thcir corresponding retention times,

correspondcd to a molecular weight of approsimately 1% D a (Figure 7) . Sodium dodecyl

su1 phate-pol yacrylamide gel electrophoresis of puri fied CC49 mA b showed onl y one protei n

band identified as having a molecular weight between 1 16-200 kDa (Lane B. Figure 8). Size-

exclusion HPLC of streptavidin-CC49 revealed two peaks w 1 th retention timcs of

approiiimatel y 7.3-7.8 minutes and 8.0-8.5 minutes, represen ting streptavidin-CC49 and

unconj ugated CC49 mA b, respective1 y (Figure 9). B y estrapolation from the standard cune

descri bed above, the molecular weigh t of streptavidin-CC49 was estimated at 200-300 lïDa.

Size-esclusion HPLC did not detect the presence of an- unconjugated streptavidin. rv hich

typically has a retention tirne of approsimately 9.0 minutes. Sodium dodecyl sulphate-

polyacqlamide pl electrophoresis of streptavidin-CC49 preparations identified three protein

bands (Lane C, Figure 8). The top two bands represent streptavidin-CC49, while the bottom

band represents unconjugated CC49 mAb. The streptavidin-CC49 bands were identified as

having a molecular weight Qûû ma, while the band representing unconjugated CC49 mAb

was identified as having a molecular weight between 116-2ûû ma. Sodium dodecyl sulphate-

I I ' I ' T S 1 . l . , , [ ,

O 2 4 6 8 1 ' 1 ' 1

10 12 14 16 18 20

Time (minutes)

Figure 7. Size-exclusion high pressure liquid chromatography analysis of CC49 mAb purified from mouse ascites fluid. The retention time of 8.24 minutes corresponds to a molecular weight of approximately 150 kDa.

Figure 8. Sodium dodecyl sulphate-polyacrylamide gel electrophoresis of CC49 mAb and streptavidin-CC49. Samples were loaded onto a 4-20% Tris-glycine gel under non- reducing conditions and identified by Coomassie Blue R-250. Lanes A and E: molecular weight standards, myosin (200 kDa), P-galactosidase ( 1 16 D a ) , phosphorylase (97 kDa), bovine serum albumin (66 kDa), ovalbumin (45 kDa), carbonic anhydrase (3 1 kDa), and lysozyme (14 kDa); Lane B: CC49 mAb. Band corresponds to a molecular weight between 116-200 kDa; Lane C: streptavidin-CC49. Top two bands represent streptavidin-CC49 and correspond to a rnolecular weight >200 kDa. Lower band represents CC49 mAb. No unconjugated streptavidin was ideatified; Lane D: streptavidin. Protein dissociated into its subunits of 15 kDa.

Time (minutes)

Figure 9. Size-exclusion hi& pressure Iiquid chromatography analysis of streptavidin- CC49. The peaks with retention times of 7.60 and 8.47 minutes represent streptavidin- CC49 and unconjugated CC49 mAb. respectively. The streptavidin-CC49 corresponds to a molecular weight between 200-300 kDa and the unconjugated CC49 rnAb corresponds to a molecular weight of approximately 150 kDa. No unconjugated streptavidin was detected. which has a retention time of approxirnately 9 minutes.

polyacrylamide gel electrophoresis confirmed those results obtained from size-exclusion HPLC

in that, unconjugated streptavidin was not present in the streptavidin-CC49 preparations. By

cornpanng the area under the peak representing unconjugated CC49 mAb o n the HPLC

chrornatograrn (Figure 9) to areas under the peak of known amounts of CC49 mAb, it was

determined that the fraction of unconjugated CC49 mAb in the streptavidin-CC49 prepantions

was typically 50%.

The biodistribution of streptavidin-CC49, labeled with i z s l , at two different tirne

intervals is summarized in Table SA. At both time intervals, the tumour, liver, and spleen

demonstnted the highest uptake of ndioactivity. Tumour uptake of radioactivi ty was 6.8 1 t

3.08% id/g d'ter the 40 hour time interval, whereas that after the 73 hour time interval was 4.16

I 1.22% idlg These values were not significantly different, however, 1251 radioactivity levels

in the blood and liver after the 40 hour time interval were significantl y higher compared to

those after the 77 hour time interval. For the remaining normal tissues, levels of- 1 2 q

ndioactivity were not significantly different after either time interval. For the streptavidin

control, the biodistnbution was evaluated at 46 hours pi in order to be comparable to the

biodistnbution of streptavidin-CC49 after the 40 hour time interval. Uptake of streptavidin-

CC49 in the tumour, liver, and spleen was approsirnately 3-, 7-, and 4-fold higher,

respectively, compared to that of the streptavidin control. Kidney uptake of streptavidin,

however, was extremely high, approsirnately 35-fold higher than that of streptavidin-CC49.

The % id of 1251 radioactivity per whole tissue was also determined and is summarizcd in Table

SB.

The T:NT ratios for streptavidin-CC49, at two different time intervals, are presented in

Table 6. For both time intervals, the T:NT ratios were r l : 1, with exception of the tumourliver

ratios. The TNT ratios for the 40 hour time interval ranged from 0.5: 1 (tumour b e r ) to 13.1 : 1

(tumour intestines) and those for the 72 hour tirne interval, m g e d frorn 0.4: 1 (tumour 1 iver) to

11.7: 1 (tumour:intestines). For the streptavidin control, the T : N ratlos were also r 1: 1. with

Table 5B. Biodistribution of strephvidin-CC49 using ü 40 or 72 hour timc iniervül between administration OS ( l *51-streptuvidin)-CC49 and ] l lin-DOTA-biotin.

% injected dose I SDt

Tissue Streptavidin-CC49t Streptavidin-CC49t Streptavidin (control) 46 ho1ir.s pi 78 koitrs pi 46 Iioctrs pi

Tumour 0.84 0.57 0.14 i- 0.09 O. 1 1 k 0.08

Blood 5.47 I 2.34 3.21 I 0.55 1.95 î 0.46

Kidneys 0.39 I O. 13 0.33 * 0.08 1.71 I 0.25

Lungs 0. 18 I 0.04 0.08 î 0.02 O. 14 k 0.03

Liver 1 2.80* 2.18 4.84 I 0.62 3.26 * 0.69

Spleen 0.57 I O. 10 0.45 =t 0.06 O. 13 * 0.02

Stomach 0.90 I O. 15 0.80 =t 0.22 0.14 I 0.02

Intestines 1.47 I 0.44 1.45 I 0.39 3.11 14.15

LS l'MT iumour-bearing mice were administercd 250 cg (50-60 [ C i ) of ( 1 251 -strcpiüvidi n)-CC49 by intraperi toneal (ip) injection followed by 1 pg ( 100 !Ki) of 1 1 11 n-DOTA-bioiin, üdrninistered by tail vein injection, üt 40 or 72 hours post-injeciion (pi) of ( 2%

streptavidin)-CC49. Mice were sücriliccd ai 6 hours pi of 1 1 'In-DOTA-biotin. Control mice were üdministei-ed 50 pg (50-60 pCi) of 1251-streptavidin by ip injection and sacrificd üt 46 hours pi. t Mean of 5-6 mice. t Biodistribution is ai 6 hours pi of l ]In-DOTA-bioiin.

Table 6. Tumour:normal iissue ratios of strepiavidin-CC49 using a 40 or 72 hour time iniervd between administration of (12"-strepiuvidin)-CC49 ünd ]ln-DOTA-biotin.

Turnour: Normal Tissue Ratios SDt

Normal Streptavidin-CC49t Streptavidin-CC49t Streptavidin (control) Tissue

46 Iioitrs pi 78 Iiorrrs pi 46 hortrs pi Blood 4.3 I 2.3 6.0 i 2.0 3.4 0.5

Kidneys 6.5 I 3.2 4.9 I 2.3 O, 1 * 0.01

Heart

Lungs

Liver 0.5 i 0.3 0.4 I O. 1 1.0 I 0.3

Spleen 1.5 i 1 .O 1 .O * 0.3 4.6 I 7.0

Stomach 6.4 I 6.8 4.8 I 3.7 5.4 I 1.6

Intestines

Femur

LS 174T iumour-bearing mice wcre adminisiered 250 pg (50-60 pCi) of (1251-strepiüvidin)-CC49 by intrapcritoneal (ip) injection followed by 1 [cg (100 pCi) of I l 'ln-DOTA-biotin, adminisiered by tail win injection, at 40 or 72 hours post-injection (pi) of (12%

streptavidin)-CC49. Mice were sacriliced üt 6 hours pi of 1 ]In-DOTA-biotin, Control mice were administered 50 pg (50-60 [ici) of 251-streptavidin by ip injection and sacri ficcd üt 46 hours pi.

f Calculated as the ratio of the 96 injected doselg of turnour and the 'Pi, injected doselg of normal tissuc. Mean of 5-6 mice. Tumour:normal tissue ratios based on biodisiribuiion ai 6 hours pi of 1 [In-DOTA-biotin.

exception of the tumour:kidneys ratio. The T: NT nt ios ranged from O. 1 : 1 (tumour hdneys) to

5.8: 1 (tumour3ntestines) (Table 6).

The biodistribution of DOTA-biotin, labeled wi th ' 1 'In, a t different time intervais is

presented in Table 7A. There was no significant dilference in the tissue upiake of radioactivi ty

Iollowing administration of DOTA-biotin at either time interval. The C/c id/g of W n

radioactivity remaining in the blood was ais0 not significantly different between the two time

intervals. However, pre-targeting with the 40 hour time interval resulted in a 5- and 26-fold

higher concentration of 1 1 I I n radioactivi ty in the tumour and blood, respective1 y, compared to

the DOTA-biotin control. The % id of 1 1 'In radioactivity per whole tissue was also calculaied

and is presented in Table 7B.

The T:NT ratios for DOTA-biotin, at two different time intervals, are presented in Table

8. For both time intewals, the T:NT ratios were > 1: 1. with exception o I the tumour:kidneys

ratios. For the 40 hour time interval, the TNT ntios ranged from 0.5: 1 ( tumourkdneys) to

19.3: 1 (tumour:stomach) and for the 77 hour time interval, the ratios ranged from 0.4: 1

(tumour.kidneys) to 17.6: 1 (tumour:stomach). For the DOTA-biotin control, the T:NT ratios

ranged from O. 1: 1 (tumourAadneys) to 14: 1 (T:BLD) (Table 8).

Forty hours was chosen as the time interva! to be employed for the pre-targeted RIT

study, and hence, was used to evaluate increasing DOTA-biotin doses of 1, 20. 40, and 150

pg (sec Discussion). As the administered dose of DOTA-biotin was increased, the absolute

tumour uptake of DOTA-biotin, expressed as pmolslg of tumour. also incrcased (see:

Appendis VI for calculations). A t doses of 1,10,40, and 150 pg, the absolute tumour uptake

of DOTA-biotin was 7.05 i 2.68, 38.47 i 5.84, 123.78 I 96.13, and 240.88 -t. 59.34

pmolslg of tumour, respective1 y. At a constant dose of 250 pg, streptavidin-CC49 achieved an

absolute tumour uptake of 34.37 I 26.78 prnolslg of tumour (see: Appendix VI). Hence,

administration of DOTA-biotin a t doses of 1, 20, 40, and 150 yg correspondcd to DOTA-

bi0tin:streptavidin-CC49 ratios per g of turnour of 0.19 I 0.06. 1.69 I 0.2 1 . 2 8 6 & 0.76, and

9.78 I 3.39 (see: Appendin VI) (Figure 10A). However, a l ter normalizing for blood

O O 'C' .-

DOTA-biotin Dose ( pg) DOTA-biotin Dose (pg)

Figure 10. l'umour uplake of DOTA-biotin ai increasing doses. 1 3 174'1' tuniour-beariiig iiiice were adiiiiiiistered 250 pg (50-60 { C i ) of ("'l-streptavidin)-CC49 by intraperitoneal injection and at 40 hours post-injection (pi) were adininistered 1, 20, 40, and 150 pg (100-300 pCi) of "'ln-DOTA-biotin by tail veiii injection. The mice were sacrificed at 6 hoiirs pi of "'ln-DOI'A-biotiii. Data points represent the mean DOTA-biotin:streptavidii~-CC49 ratio per g of tuciiour f SD (11=6 exçept at the 1 pg dose, ~ 5 ) . (A) Ratios are not normalized to "'ln radioactivity in blood; (H) Ratios are normnlized to " ' I n radioactivity i i i hlood. (sec: Appendix VI for culculation of ratios).

radioactivity, the DOTA-biotin:strepla\.idin-CC49 ratios per g of tumour were 0.70 + 0.19.

0.99 +. 0.22, 1.33 i 0.41, and 1.94 I 0.46 at DOTA-biotin doses of 1. 10, 40, and 150 pg ,

respectively (sec Appendii VI) (Figure IOB). Figure 10A shows that without norrnalization

for blood radioactivity, the DOTA-biotin:streptavidin-CC49 ratios per g of tumour increased

almost linearl y with increasing doses of DOTA-biotin. In contrat. Figure 1 OB demonsirates

rhat with norrnalization for b l d radioactivity, the ratios reached a plateau as the DOTA-biotin

dose Kas increased from 40 to 150 ug.

Discussion

To i n c r w e the l ike l ihd of successful pre-targeted RIT. an appropriate dosing

schedule musr be ernployed. For 2-step approaches. sufficien t amounts of mA b conj ugate

should be administered to saturate the availabie binding sites in the tumour. Secondl!-, the time

inten-al should be one that allows for sufficient elimination of the anti body [rom the blmd and

normal tissues to occur. and in addition. one that maintains as hiph a level of antibody in the

tumour as possible. Lastiy. the administered dose of effector molecule shouId be one that

saturates the a\*ailable hapten binding sites in the tumour to deliver as high a radiation dose to

the turnour as possible. Dual label biodistiibution studies were performed in LS 174T tumour-

beanng mice to esperimentally determine a Ume inten-ai and DOTA-biotin dose that meet these

cnterk Two different time intends (30 and 72 hours) and four different doses of DOTA-

biotin ( 1. 70,4û. and 150 pg j were evaluated

The T:Nï ratios for streptavidin-CC@. at both time inten-als, were 2 1 : 1. u-i th

exception of the tumour:liver ratios (Table 6), suggesting that tumour locali~ation wirh

streptavidin-CC49 was achieved. The low tumour.liver ( ~ 0 . 5 : 1 ) and tumour spleen ratios

( 6 1 -5: 1 ), at both tirne intenais, are a result of the high liver and spleen uptake of radioactivi ty.

The lirer and spleen are reported to be responsible for the catabolism of antibodies and oiher

protein macromolecules (Wagie et al.. 1978), rnost likel y esplaining the relative1 y high uptake

of radioactivity demonstrateci bv these tissues. Liver uptake of radioactivity, ai both time

intervals, was 10.37- 15.99 %id& and spleen uptake was 4.13-4.904 i#g (Table 5). For the

streptavidin control, however, liver uptake of radioactivi ty was on1 y 7.17 I 0.46% id/g and

spleen uptake was only 1.24 I 0.14% idg. The apparent higher u p a e of streptavidin-CC49

in the liver and spleen ma- be due to Fc receptor-mediated uptake of the mAb conjugate (via the

Fc regions of the CC49 mAb) by the Kupffer cells, hepatocytes, and other Fc receptor-beanng

cells residing in these tissues. With exception of the tumou~hdneys ratio. the TNT ratios for

the streptavidin control were r 1 : 1, suggesting that the protein achieved some non-specific

turnour localization (Table 6). Similar non-specific tumour uptake of streptavidin has also k e n

reported by others. Hnatorvich et al. ( 1993) successfully imaged LS 174T colorectal senogdts

by admi nistering unconjugated streptavidin followed by 1 1 [In-EDTA-biotin. Zhang et al.

( 1997) administered 1311 -avidin alone to treat SH 1 N3 human ovanan cancer xenografts,

resulting in a significant reduction in tumour ueight. Thomas et al. ( 1989) suggested that

tumour capillaries may be up to IO-fold more permeable than capillaries of normal tissues.

possi bl y esplaining the non-specific tumour uptalie of streptavidin obsen-ed in this and other

studies. Very high kidney uptake of radioactivity (38.96 I 7.809 id/@ following

administration of the streptavidin control nras obsenPed in this study (Table 5) . Similar

obsen.ations in mice biodistri bution studies have also been reported by others. Hnaiowich et

al. ( 1993) obsen-ed a hdney uptake of streptavidin of 759 id&, while Schechter et al. ( 1990)

obsewed hdney uptake to be as high as 804 idlg. Schechter et al. ( 1990) proposed that the

high kidney uptake of streptavidin may have to do with the presence of truncated forms of

streptavidin in the preparations. Commercial preparations of streptavidin have k e n reported to

contain truncated forms of streptavidin, generated as a result of certain isolation procedures

perfomed during production. During these procedures, streptavidin apparently can undergo

proteolytic degradation. resulting in the cleavage of 12- 14 amino acid residues at the N-

terminus and a maximum of 18 residues at the C-terminus (Bayer et al., 1986). Although not

completely understd, it is hypothesized that the e x p e d arnino acid residues resulting from

truncation may interact wi th tissues like the kidneys. consequently resulting in high kidney

uptake of streptavidin (Schechter et al. 1990). A commercial preparation of streptavidin \vas

used in this study, as well as in the studies performed by Hnatowich e t al. (1993) and

Schechter, e t al. ( 1990). However, the streptavidin preparations came from different

manufacturers in al1 three studies. Interestingly, when streptavidin was conjugated to the CC49

mAb. the hdney uptake of raciioactivity was 35-fold lower. This suggests that by bang

conjugated to the mAb, streptavidin was sornehow prevented from interacting with portions of

the hdneys that, othenvise, would mediate uptake of the streptavidin protein.

The T:NT ratios for DOTA-biotin. at both time intemals, were r 1 : 1. wi th esception of

the nirnouckidneys ratios (Table 8). suggesting that the 3-step pre-targeting approach achieved

tumour lowlization of DOTA-biotin. The turnourdadneys ratios for both time i n t e n d s were

10.5: 1 and are a result of the relativelu h g h uptake of radioactivity by the kidneys. For both

time intenais, the hdnevs represented the highest uptake of radioactivi ty a t 1.26-2.10% id&,

possi bl y reflecting the rend excretion of DOTA-biotin. The T:NT ratios for the DOTA-biotin

control suggesi that it achieved some non-specific tumour localization (Table 8). For example.

the T:BL ratio for the DOTA-biotin control was 14: 1. which is i f o l d higher than that of

DOTA-biotin following pre-targeting rvith the LU) hour tirne interval. However, this is largely

because b l d levels of radioactivity rollowing administration of the DOTA-biotin control were

estremely loa. (0.01 I 0.003% id/@ (Table 7). Tumour uptalre (Q id/@ of the DOTA-biotin

control \vas actuall y I fo ld less than that of DOTA-bioti n following pre-targeting wi th the 10

hour time interval.

After the 72 hour time interval. levels of 1'51 radioactivity were significan* lower in

the blood and liver, suggesting that the longer time i n t e n d allowed for greater elimination of

streptavidin-CC49 from these tissues to occur (Table 5). For the rernaining normal tissues and

turnour, however, loaer tevels of streptavidin-CC49 &ter the 72 hour time inten-al emerged

only as a trend. This trend suggests that the 72 hour time intenal ma). have allowed for greater

elimination of antibody from these normal tissues to occur but, i t ma) have also resulted in less

turnour- bound strepm-idin-CC49 available [or binding to DOTA- biotin. No time interval

shorter than 40 houn was investigated, and therefore, the extent of antibody elimination from

the blood and normal tissues achieved over the first 40 hours pi is unknown. Nevertheless, to

deliver as much goy-DOTA-biotin and as high a radiation dose to the tumour as possible, 40

hours appears to be the appropriate tirne intewd for the pre-targeted RIT study.

Figure 10A demonstrates that administering higher doses of DOTA- biotin i ncreased

DOTA-biotin uptake in the tumour. as is evident by the increasing DOTA-biotin:streptavidin-

CC49 ratios per g of tumour. By usinp dual label techniques. we were able to m w u r e the

DOTA-biotin:streptavidin-CC49 ratios per g of tumour, which in turn, allowed us to determine

which dose of DOTA-biotin achieved saturation of the biotin binding sites in the tumour.

Without nomalization for blood radioactivitg, the DOTA-bi0tin:streptavidin-CC49 ratios per g

of tumour increased aimost linearly with the DOTA-biotin dose (Figure 10A). For the 1 pg

dose, the mean ratio was 0.19 I 0.06 and for the 150 !cg dose, 9.28 * 339. In theory, these

ratios should reach a plateau with increasing dose because of the finite amount of biotin binding

sites available in the tumour due to using a f k e d dose of streptavidin-CC49 (250 [cg).

Theoreticaily, if conjugation of one streptavidin molecule per CC49 mAb was achieved, the

maximum DmA-biotin:streptavidin ratios per g of tumour achievable would be 4: 1 due to the

tetravalent binding between biotin and streptavidin. On the other hand. if conjugation or two

molecules of streptavidin per CC49 mAb was achieved. the maximum ratios achievable would

bc 8: 1. Eitber scenario is possible since the molecular weight of streptavidin-CC49 uas

estimated bp SDS-PAGE (Lane C. Figure 8) and size-exclusion HPLC (Figure 9) to be

between 300-300 ma, ie. IgG mAbs have a molecular weight of 1M kDa. while streptavidin

has weight of 65 D a .

lncreasing the DOTA-biotin dose would aiso increase the arnount of DOTA-biotin

present in the blood at the time of anaiysis. either as free DOTA-biotin o r DOTA-biotin

complesed wi th circulating streptavidin-CC49. To correct for the contri bu tion of radioactive

DOTA-biotin in the blood circulating through the tumour, the absolute tumour uptake of

DOTA-biotin (pmoldg of tumour) was normalized to the absolute arnount of DOTA-biotin in

blood (pmoldg of b l d ) (see: Appendix VI for calculations). Similarly, the tumour uptake of

streptavidi n-CC49 was normaiized to the amount of streptavidi n-CC49 in the blood. When

corrected for blood radioactivity, the DOTA-biotin:strepiavidin-CC49 ratios per g of tumour no

longer increased linearly with the DOTA-biotin dose but rather, reached a plateau as the dose

was increased from 40 to 1 9 pg (Figure IOB). The maximum ratio achieved at the 150 pg

dose was 1.94: 1. Pharmacolunetic modeling by van Osdol et al. (1993), suggested that

equilibration of the biotin binding sites in a turnour nodule, pre-targeted with streptavidinylated

mAb, could be reached within 4 hours of administration of biotin. The pIateau observed with

increasing doses of DOTA-biotin suggests that saturation of the biotin binding sites in the

LS 174T xenografts was achieved, both within the 6 hours elapsed since administration or

DOTA-biotin and within the range of the DOTA-biotin doses administered. The DOTA-

biotin:streptavidin-CC49 ratios pet- g of tumour (with normalization for b l d radioactivity),

obtained from increasing doses of DOTA-biotin, were significantly different [rom each other

(p4.ûûûL). I t appears that the most effective dose for the pre-targeted RIT study is 40 pg.

This dose approached near saturation of the biotin binding sites, and in addition, will allow for

DOTA-biotin to be administered at a higher specific activity, ie. pCi/pg, than will a dose or 150

pg. Higher specific activi ty is desirable since on1 y a fini te amount of DOTA-biotin can bind to

the turnour-bound streptavidin-CC49. Turnour uptake of DOTA-biotin radiolabeled at higher

specific activi ties could increase the radiation dose delivered to the tumour.

Our results have shown that tumour Iocalization of DOTA-biotin wi th pre-targeting by

streptavidin-CC49 was achieved. Moreover, we developed a dosing schedule for the pre-

targeted RIT study, which involves administration of 250 pg of strephvidin-CC49, followed

by a 40 hour time interval, and 40 pg of gOY-DOTA-biotin. The time intend of 40 hours and

DOTA-biotin dose of 40 pg were chosen because they met specific cnteria Compared to the 77

hour time interval, the 40 hour interval demonstrated a trend towards higher tumour levels of

streptavidin-CC49. The 40 pg dose of DOTA-biotin appears to have been sufficient for

saturating the biotin binding sites in the tumour. However, whether or not these parameters

will result in RIT success on only be answered after the pre-targeted RIT study is c m i e d out.

EXPERIIMENT III:

PRE-TARGETED RADIOIMMUNOTHERAPY WITH STREPTAVIDIN-CC49 MONOCLONAL ANTIBODY

AND Bq-DOTA-BIOTIN

Introduction

Re-targeted RIT is an approach designed to be less toxic to bone marrow than

conventionai RIT with radiolabeled antibodies. Pre-clinicai (Axworthy et al., 1995) and clinid

(Cremonesi et al., 1997; Paganelli et al.. 1993) investigations of pre-targeted RIT with the

(strept)avidin-biotin system have y ielded promising resul ts. For esample, Axworth y et al.

(1995) dministered up to 800 yCi of goY-DOTA-biotin to tumour-bearing mice following pre-

targeting with streptavidinylated mAb. Complete tumour regressions were seen in 81 10 treated

mice bearing human breast tumours and in 10/10 treated mice bearing human colorectal and

small ce11 lung carcinoma turnours. Dosimetry estimates predicted that tumour doses of > 150

Gy1800 pCi were achieved. No apparent toxicity was observed with pre-targedng, however,

conventional treatment with 400 pCi of -mAb was associated with 100% lethality in al1

mice and was less therapeutically effective than pre-targeted RIT. In a clinical study by

Paganelli et al. ( 1993). five patients were treated with 40-M mCi of 9OY -DOTA-biotin in a 3-

step pre-targeting approach and demonstrated nedi gi ble hematological tosici ty . Subsequentl y,

415 patients received a second course of treatment under a sîmilar protocol wi th 100 mCi of

9OY-DOTA-biotin. Estimated bone marrow doses were 0.07 and 0.13 Gy for the first and

second treatments, respectively. One patient demonstrated a minor response, one had stable

disease. and three had progressive disease.

The objective of this esperiment was to evaluate the thenpeutic efficacy and saî-ety o l

pre-targeted RIT with streptavidin-CC49 and -DOTA-biotin, using the dosing schcdule

previousl y developed in Expriment II. The dosing schedule involves ip administration of

streptavidin-CC49 at a dose of 7 3 pg, followed by iv administration of 40 yg of 90Y-M3T~-

biotin 40 hours later. The 40 hour time interval was chosen because it demonstrated a trend

towards greater turnour retention of streptavidin-CC49 cornpared to the 79 hour time interval.

The 40 pg dose of DOTA-biotin appeared to be the optimal dose for saturating the hapten

bindinp sites in the tumour and for delivenng as high a radiation dose as possible to the tumour

cornpared to other DOTA-biotin doses of 1, 20, and 150 p g Radiolabeled CC49 mAb has

previously been shown in animal models to be effective in treating tumours. A I f o l d reduction

in tumour volume was observed following the administration of 300 pCi of 1311-CC49 mAb to

mice bearing HT-39 xenografts (Greiner et al., 1993). In a subsequent study, Greiner e t al.

(1994) treated E-IT-79 turnour-bearing mice with multiple doses ( Ix ) of 300 pCi of l 3 %CC49

mAb and drarnatically inhibited tumour growth for the first 21 days after treatment. Not

surprisingly, clinical RIT with CC49 mAb has shown less promising results than animal

studies, as has clinicai RIT of most solid tumours. In a Phase 1 trial, twenty-four patients were

treated with 13'1-CC49 mAb at doses of 15-90 mCilm2(Divgi et al., 1995). Only stable disease

and one minor response were observed, with hematological toxicity appearing at doses >60

mCilm2. Using the dosing schedule developed in Experiment I I , athymic mice bearing

subcutaneous human LS 174T xenografts were treated wi th 900 pCi of 90Y-DOTA-biotin as

part of a small pilot study. Tumour growth, peripheral white blood ceil (WBC) counts, and

whole body weights were evaluated over a p e n d of 25 days following the administration of

9oY -DOTA-biotin to assess the thenpeutic efficacy and toxici ty of the pre-targeted RIT

regimen.

Methods

Preparaiiotr oJS~repfa vidin - CC49 MotroclotraI An fibody

CC49 mAb was purified h o m mouse ascites Iluid and analyzed for purity by s i x -

csclusion HPLC and SDS-PAGE (see: Appendix 1). CC49 mA b was then conjugated to

streptavidin and the conjugate evaluated for purity and molecular weight by size-exclusion

HPLC and SDS-PAGE (see: Appendis II). The immunoreactive fraction of streptavidin-CC49

was 6 3 6 (see: A ppendix III). Pnor to administration, s treptavidin-CC49 was steri lized by

passing the antibody solution through a 0.22 pm HT Tuffryn membrane syringe fil ter (Gelman

Sciences, Ann Arbor, MI) and then tested for sterili ty by the USP XXIII sterili ty test. Tests

rcsults were negative for bacterial and fungus contamination.

Prepration of 9OY-DOTA-biotin (see: A ppendix 1 V)

DOTA-biotin was labeled with chloride to a specific activity of approxirnately 22-

23 pCi/pg. T h e radiochernical pun ty of -DOTA-biotin assessed by reversed-phase

chromatography was 99% and the streptavidin- binding capaci ty was 92%. T o mi ni mize

radioIysis of DCrTA-biotin, ascorbic acid (20 mg/mL) (Fisher Scientific, Markham, Ont.) was

added, DTPA calcium trisodium salt hydrate ( 10 mM) (Fluka Chemika, Switzerland) was also

added to chelate any free 90Y ions present in the preparation. Yttrium 90-DOTA-biotin was

diluted for injection with sterile 0.9% saline and the final solution was sterilized as prcviousl y

described. Yttrium 90-DOTA-biotin was aiso tested for sterility by the USP XXIIl stenli ty test

and was negative for bacteriai and fungal c o n ~ i n a t i o n .

Radioirnmroiothernpy with StrepiavidimCC49 Monocfomf Ar1 tibody ami 9oY-DOTA-bio~iri

LS 174T h u m a . colon cancer cells were injected subcutaneously in twelve six- to eight-

week old femaie athymic mice and the tumours allowed t o grow for 12 days, at which time

tumour diameters ranged from 3-6 mm (see: Appendis V). One group of tumour-bearing mice

(n=6) received Z M k ~ g of streptavidin-CC49 by ip injection and 40 hours later received 900

pCi (40 pg) of goy-DOTA-biotin by tail vein injection. Prior to performing tail vein injections

of 90Y-DOTA-biotin, d l mice were anesthetized as previously descnbed in Experiment 1. The

rcmaining tumour-bearing mice served as the control group (n=6) and received no treatrnent.

Turnour volumes in both groups of mice were measured 2-3 times a week for 15 days

Sollowing treatrnent with gOY -DOTA-biotin, and were calculated using the Sollowing formula:

volume (mm3) = 4/3x[(rl +rz) + 213, where rl and r2 are the radii of the long and short axes OS

the tumour rneasured by precision calipers. Bone marrow toxicity was evaluated by mûasuring

pet-ipheral WBC in al1 mice ? o r 3 times a week for 25 days following treatment. Periphenl

WBC counts were perfonned by withdrawing whole blood via the tail vein using 47.7 PL-

sized heparinized rnicrocapillary tubes. Prior to the collection of blood samples, al1 mice werc

anesthetized as previously descri bed in Experiment 1. Blood samples were diluted 1 :?O: vlv

with Turk's solution containing 3% acetic acid and 0.01% Crystal Violet stain (Sigma

Chernical Co.) and incubated for at Ieast 25 minutes at room temperature. White blood celis

were then counted on a standard hemocytometer. Tolcicity was also monitored by measurine

the whole body weight of each m o w 7 or 3 times a week for 25 days follouing treatment.

Throughout the entire study. the mice uere housed in stenle cages fitted with filtered lids and

maintained on a diet of stenle water and rodent chou (Autocla~~able Rodent Chou, 50 10, PMI

Feeds, St. Louis, MO).

S fatistical Arzalysis

The mean tumour volume was compared between groups using the Student's unpaired

t-test with a level of significance of p4.05.

Results

Size-exclusion HPLC of CC49 mAb purified from moue ascites fluid showed a single

peak with retention tirnes of 8.3-8.6 minutes. Sodium dodecyl sulphate-polyaqlamide pl

electrophoresis of purified CC49 mAb shoued only one protein band identified as having a

rnolecular weight between 1 16-20 kDa Size-exclusion HPLC of streptavidin-CC49 revealed

ru.0 peaks with retention tirnes of 7.3-7.6 minutes and 8.3-8.3 minutes. representing

streptavidin-CC49 and unconjugated CC49 mA b, respective1 y. Two bands representing

streptavidin-CC49 and unconjugated CC49 mA b correspondi ng to molecular wei ghts 900

kDa and between 1 16-200 kDa, respectively, were identified by SDS-PAGE. Both size-

escl usion HPLC and SDS-PAGE did not detect the presence of an). unconj ugated streptavidi n.

Anal ysis of the size-escl usion HPLC chromatograms of streptavidin-CC49, as previousl y

described in Esperiment II, re\*eaied that approsimately 9% of CC49 mAb presen t in the

streptavidin-CC49 preparation remained unconjugated.

At the time of streptavidin-CC49 adminktration tumour volumes rangeci from 8 7 6 8

mm3 (mean=81 mm3) and from 4& 113 mm3 (mean=74 mm3) for mice in the untreated and

treated groups, respectively. One mouse from the untreated group was sacrificed on the

seventh day following treatment with goy-DOTA-biotin, due to significant ulceration of the

tumour. On days I I and 14 one mouse from the treated group and one mouse from the

untreated group, respectivelv. died from anesthetic overdose. Another death, also attributed to

anesthetic overdose, occurred in the treated group o n day 32.

Figure 11A sho\rs individual tumour growth cun-es for al1 mice measured over the 25

da? period following treatment n-i th 9OY-DOTA-biotin, enpressed as the % change in turnour

volume. The % change in tumour volume uas calculated as: [(tumeur volumed,, - tumour

~o lu rne~ , ,~ ) + turnour ~ o l u r n e ~ , ~ , ] .u 100%. O n e mouse (#3) from the treated group

demonstrated a noticeable response to therapy. After an initial increase in tumour vol urne of

approsirnately 40% during the f in t 3 days follorving treatment, no further measurable tumour

growth was obsen-ed until 1 1 days later on day 14. On day 16 the turnour appeared to regress.

reaching a volume on day ?O that uas approsimately WC smaller than its volume pnor to

DOTA-biotin administration, te. on day 0. However, on day 12 tumour growth resumed, and

at the end of the study on day 25. the tumour volume was 1 10% larger than that measured on

day O. Figure 12 shows the anterior views of an untreated mouse (#6) and the treated mouse

descri bed above (#3) taken on day 75. On day O the tumour volumes were 1 8 0 mm3 and 144

mm3 for mouse #3 and mouse #6. respectively. However. on day 25 the tumour volume 01-

mouse #6 was approsimately 1680 mm3, whereas the turnour volume of mouse #3 treatcd nith

streptavidin-CC49 and 900 pCi of goY-DOTA-biotin a m nearly 6-fold smaller at 268 mm3.

The tumour volume of mouse #6 increased by 1000% from day 0. nearly IO-fold greater than

the change in tumour volume obsen-ed with mouse #3. Tumour growth rates of the remaini ng

treated mice uere similar to most of those of the untreated mice (Figure 1 1 A). Figure 1 1 B

shows the mean C/c change in tumour volume over time for both groups of mice, escluding al1

the mice that died or were sacrificed pnor to the completion of the study. At the end of the

study on day 25. the rnean R change in tumour volume nas 1974 r 1475% (n=4) and 692 t

410%~ (n=4) for the untreated and treated mice, respective1 y. These values were not statisticall y

different. Tumour volumes are also presented in Table 9.

The mean peripheral WBC counts for either group was highly variable throughout the

35 &y evaluation period but, there appeared to be no overall decline in WBC counts in the

treated mice that would indicate WBC depression 13A). Whole body weights for both

groups of mice varied to a lesser extent than the mean WBC counts (Figure 13B). The mean

body weights of treated and untreated mice also did not decline dunng the 25 day period,

indicating no apparent deterioration of heaith in the animals.

Discussion

Pre-targeted RIT is an approach which, compared to conventional RIT with

radiolabeled mAbs, may reduce bone marrow toxicity and possibly allow higher amounts of

radioactivity to be safely administered to the patient. Several pre-clinical and clinical

investigations of pre-targeted RIT wi th the (strept)auidin-biotin system have been performed

and have yielded encouraging results (Axworthy et al., 1995; Cremonesi et al., 1997: Paganelli

et al., 1993). In this small pilot study, we evaiuated the therapeutic efficacy and tosicity of

streptavidin-CC49 and goy-DOTA-biotin in athymic mice bearing subcutaneous human

LS 174T colorectal senografts. We administered 150 pg of streptavidin-CC49 by i p injection

and 40 hours later, administered 900 pCi (40 pg) of g%DOT~-biot in by iv injection.

Folloaing treatrnent with 9OY -DOTA-biotin, tumour growth. pet-ipheral WBC counrs. and

whole body weights uere evaluated over a 75 day p e n d following the administration of 9 V -

DmA-biot in to evaluate the therapeutic efficacy and toxicity of the pre-targeted RIT approach.

It has been postuiated that a radiation dose of at least 50-80 Gy m u t be delivered to

the tumour within a period of 1 week for successful treatment (Bruland, 1995; Ellis, 1%8).

Only 1/6 treated mice demonstrated a major response to therapy (Figure 1 1A; Figure 12).

possibly indicating that cytotoxic doses of radiation may not have been delivered to al1 the

tumours treated with the streptavidin-CC49 and -DOTA-biotin regimen. Nevertheless, at

Table 9. Individual tumour \duines of LS 174T senogrdis Ibllo\ving treatment wi th strcpiiividin-CC49 and ")Y -DOTA-biotin.

Tumour Volume (mm31

Post-treatment Untreated Mice 1 O Y 3 4 5 6

- 1.6t 87 368 8 14 34 76 01 180 449 87 48 1 0 1 44 3 383 850 56 113 161 382 5 449 905 351 32 1 180 534 7 606 1023 651 368 268 697 9 849 796 323 322 796 1 1 1085 0 5 449 449 1033 14 1288* 1085 524 534 1 15û 16 12% 1 564 128û 18 1288 1515 3% 1515 20 15% 1515 7% 15% 23 1857 1857 7% 1515 25 1950 1950 796 1680

Treated Mice 1 2O 3 4 5 6

LS l74T tumour-bearing mice wcre: i ) treated wi th 2 M kig of strcptavidin-CC49 by intriperi toneal injection and 40 hours later were administered 900 pCi (40 !tg) of ")Y -DOTA-biotin by tail vein injection or i i ) lell untreüted. t Strepiiividin-CC49 administered. $ m~ -DOTA-biotin administered. Y Mouse sacrificeci on day 7 due t« signilïcant ulccration of the tumoiir.

Mice died from anesthetic overdose on düys indicated by "*".

-0- Treated

T n

a b

O . . . . l . . W . , . l . . l . . . l I , , , , I I , l l , , l , , ,

+ + -5 O 5 10 15 20 25 30

Days Post-treatment

l -+- Untreated * Treated l

5: a b

0- ~ ~ ~ l , ~ . ~ ~ , l , , , , , , l , l , , , , l , l , , I , , , , ,

+ + -5 O 5 10 15 20 25 30

Days Post-treatment

Figure 13. Evaluation of 'Y-DOTA-biotin toxicity following pre-targetcd radioinimunotlierapy. LS 1741. tuniour-bearing mice were: i ) treated with 250 pg of streptavidin-CC49 by intraperitoneal injection and 40 hours later were administered 900 pCi (40 pg) of 90Y- DOTA-biotin by tail vein injection (open circles) or i i ) leA unireated (closed circles). (A) Mice were evaluated for myelosuppression by monitoring petipheral white blood cell counts; (11) Whole body toxicity was evaluated by monitoring whole body weights. Data points represent mean values * SD (n=4 mice/group). All mice that died or were sacrificed prior to the cornplation of the study were excluded. B Arrow denotes admiiiistration of streptavidiii-CC49.

Arrow denotes administration of MY-DOTA-biotiii.

the completion of the 25 day evaluation period, the mean 8 change in tumour volume for the

untreated mice was 1974 * 1475% (n=4), whereas that of the treated mice was 692 i 410%

(n=4) (Figure 1 1B). These values were not significantly different, however, the trend suggests

that the treatment regimen rnay have had a therapeutic effect. Perhaps wi th larger sarnple sizes,

a significant difference may be observed. The lack of apparent toxicity a t the dose of

DOTA-biotin administered (900 yCi) may allow for larger doses of 9oY-DmA-biotin to be

d e l y adrninistered in order to achieve a higher response rate. Al temativel y, a second course of

treatment with streptavidin-CC49 and goy-DOTA-biotin 1-2 weeks following the initial course

of therapy could be given. One concem with this approach, however, is the possibility of

HAMA and HASA responses in patients, which may diminish the effectiveness of the

treatrnent-

The mean peripheral WBC counts for both groups of mice were highly variable,

possibly due to experimental error (Figure 13A). There was no obvious decline in the mean

peripheral WBC counts in the treated mice compared to pre-treatment values but, perhaps with

larger sample sizes the variability would be less and the changes in WBC counts more easily

discemible. Based on the mean peripheral WBC counts, there was no apparent bone manow

toxicity following the administration of 900 pCi o f 90Y-DOTA-biotin. I n mice,

myelosupression has typically been observed within 14 days from the time of treatment with

radiolabeled mAb (Lee et al., 1990; Senekowitsch et al., 1989; Buchegger et al., 1990).

Therelore, the length of the evaluation period (25 days) should have been sufficient to observe

myelosuppression. Mean body weight rneasurernents were not as variable as the mean

peripherd W C counts and, it was quite apparent that no loss in body weight occurred in the

treated mice, suggesting that the health of the animals did not deteriorate (Figure 13B). If

myelosuppression did occur but, was undetected, the fact that the mice were still able to thnve

following the administration of 9OY-DOTA-biotin suggests that bone marrow toxici ty, if any,

was not life-threatening. No cornpanson was made between the therapeutic efficacy and

toxicity of pre-taqeted RIT to that of conventional RIT wi th radiolabeled mAbs in this study.

However. cornparisons with other studies support the hypothesis that pre-targeted RIT may be

less toxic to bone marrow than conventional RIT. Lee et al. ( 1990), observed a IO-fold

decrease in WBC counts in mice within 2 weeks of administering only 150-750 yCi of 90Y-

CO 1% 1A mAb. S harkey et ai. ( 1988) observed as much as a 9435% decrease in WBC counts

in mice within 1-3 weeks following therapy with ?-O and 50 pCi of 90Y-NP2 mAb. Results

from a pre-clinical study by Axworthy e t al. (1995) further support Our findings and the

hypothesis that pre-targeted RIT may be less tonic to bone rnarrow. These authors observed no

apparent bone marrow toxicity in turnour-bearing mice treated with 200-800 pCi of 90Y-

DOTA-biotin after pre-targeting wi th streptavidinylated mAb. Conventional RIT wi th 400 pCi

of -mAb, however, resulted in 100% lethality to the animals. Moreover, the conventional

RIT approach was not as therapeuticall y effective as the pre-targeted approach. w hich achieved

complete turnour regressions in as high as 10J10 mice bearing colorectal and small ce11 lung

carcinoma xenografts. In this study, pre-targeted RIT of LS 174T colorectal xenografts was

also not as therapeutically effective as the pre-targeted RIT perfomed by Axwonhy e t al.

( 1995). The dosing regimen employed in this study may not necessaril y be the optimal regi men

for pre-targeted RIT in this animal model, and therefore, it rnay be essential to investigate and

compare the thenpeutic efficacy and toxicity of other dosing regimens, ie. those with a shorter

time interval or different doses of antibody and effector molecule.

Results from this small pilot study are encouraging because of the considerable

suppression in tumour growth achieved in at least one animal. Despite the relativelp high dose

of gOY-DOTA- biotin adrninistered (900 pCi), the treated mice continued to thrive and no

apparent bone marrow toxicity was observed. Furthemore, whiie only 116 mice treated with

streptavidin-CC49 and -DOTA- biotin demonstrated a considerable response to thenpy , the

observed trend towards a smaller % change in tumour volume following treatment suggests

that the treatment regimen may have had a therapeutic effect. However, larger sarnples sizes

will be required in future evaluations to confirm the therapeutic efficacy and bone marrow

toxicity of this or any similar pre-targeted RIT dosing regimen. Possibilities for increasing the

response rate include administering higher doses of goY-DOTA-biotin (S00 @i), perfonning

multiple courses of treatment with streptavidin-CC49 and g%DOTA-biotin, and using other

dosing regimens. Nevertheless, pre-targeted RIT with streptavidin-CC49 and 9 m ( - m A -

biotin has demonstrated some promise as an effective therapeutic approach to the treatment of

colon cancer but, additional pre-clinical studies are necessary (see: Conclusions and Future

Perspectives), More its dinical efficacy and safety on be evaluated in patients.

CONCLUSIONS AND FUTURE PERSPECTIVES

Two major limitations of RIT are the sub-therapeutic radiation doses delivered to solid

tumours and the bone marrow toxicity in the form of myelosuppression. Reducing bone

rnarrow tosicity is essential in improving the cliniul success of RIT because it may dso permit

higher amounts of radioactivity to be safely administered to the patient, which in turn. ma-

improve the radiation doses delivered to the tumour. We investigated a novel RIT approach

dcsigned to reduce bone marrow tonicity. which involves pre-targeting of the lesion with a

streptavidinylated mAb followed by administration of radiolabeled biotin. The overall objective

of our study was to investigate pre-targeted RIT with streptavidin-CC49 mAb and 9 0 Y - m A -

biotin in athymic mice bearing subcutaneous human LS IWT xenografts.

In Experiment 1, we demonstrated that DOTA-biotin is suitable for pre-targeting

strategies. Blood levels of DOTA-biotin, labeled with 1 Wn, in BALBlc mice were 4% id/@ at

1 hour pi (Figure 5). Blood levels of radioactivity decreased rapidly by 5- and 100-fold at 2

and 12 hours pi, respective1 y. Kidneys demonstrated the highest uptake of radioactivity at

3.39 idfp at 1 hour pi. which decreased by 8-fold to 0.4% idlg at 12 hours pi (Figure 6). This

pattern of uptiùce and elimination of radioactivity in the kidneys suggested that DOTA-biotin

was eliminated by rend escretion. In the remaining normal tissues, uptake of radioactivity was

very low and \vas measured at 4 . 3 % idlp at 3 hours pi (Figure 6).

In Esperimeni II , a dosing schedule for pre-targeted RIT with streptavidin-CC49 and

'OY -DOTA-biotin was developed. Dual label biodistribution studies in LS 174T tumour-bearing

mice allowed us to determine the time interval and DOTA-biotin dose to be used in the RIT

study. The 40 hour time interval demonstrated a trend towards greater tumour retention of

sueptavidin-CC49 compared to the 72 hour tirne interval (6.81 t 3.08% idlg vs. 4.76 1 .??

5% id/@, and hence. was chosen as the time interval for the pre-iargeted RIT study (Table 5).

When normalized for blood radioactivity, the DOTA-biotin dose of 40 eg appeared to have

reached near saturation of the biotin binding sites in the tumour, achieving a DOTA-

bi0tin:streptavidin-CC49 ratio per g of tumour of 1.33: 1 (Figure IOB). Therefore, the 40 pg

dose was chosen as the dose for the pre-targeted RIT study.

A small pilot study was carried out in Evperiment III evaluating the thenpeutic efficacy

and toxicity of pre-targeted RIT with streptavidin-CC49 and "Y-DOTA-biotin. Using the

dosing schedule developed in Expenment II, athymic mice bearing subcutaneous human

LS 174T xenografts were treated with 900 pCi of gq-DOTA-biotin. Only 116 treated mice

demonstrated a considenble suppression in tumour growth (Figure 1 1A ; Figure 12). In this

mouse, no further measurable tumour growth was observed over a period of I l days. The

tumour did appear to regress but, shortly thereafter. tumour growth resumed. The tumour

growth rates of the remaining treated mice were similar to those of most of the untreated rnice,

however, the trend towards a smaller % change in tumour volume following treatrnent

suggested that the treatment regimen may have had a thenpeutic effect. At the completion of the

study, the mean % change in tumour volume for the treated mice was 6 10 i: 410% (n=4) and

that of the untreated mice was 1974 + 1475% (n=4) (Figure 1 lB). At the relatively high dose

of 90Y-DOTA-biotin administered (900 pCi), no apparent bone marrow toxicity or

deterioration in health of the animals were observed (Figure 13).

Our resul ts demons trated that streptavidin-CC49 and 9OY -DOTA-bioti n have somc

promise as an effective RIT approach to treat colorectal cancer. Despi te on1 y observing one

considenble response to thenpy, the lack of apparent bone marrow toxicity and deteriontion

in health suggested that pre-targeted RIT with the (strept)avidin-biotin system may be less tosic

to bone marrow than conventionai RIT approaches. The pre-iargeting approach, therefore, may

aiso permit higher arnounts of ndioactivity to be administered to a patient, and perhaps.

improve the radiation dose delivered to the tumour. Hence, pre-targeting has the potential to

achieve greater cl i nicd success than conventional RIT approaches.

Utilizing other mAbs can further estend our dosing regimen beyond treati ng colorectal

cancer. Alternatively, CC49 mAb, a pan-carcinoma antibody, can potentially be used to treat

other adenocarcinornas such as, ovarian and breast cancer. The pre-meeting approach is more

complicated than conventional RIT with radiolabeled mAbs because of time intervals to

consider and the additional reagents for which optimal doses must be determined. The dosing

regimen employed in this study rnay not have necessarily been the optimal regimen for treating

LS174T xenografts. Future perspectives. therefore. for pre-targeted RIT with the

(strept)avidin-biotin systern may include further pre-clinicai evduation of this approach using

di fferent dosing regirnens, ie. shorter tirne intervals, di fferent doses of antibody , or mu1 tiple

treatment protocols.

REFERENCES

Allison MA. Jones SE. McGuffey P. et al. Phase II trial of outpatient interleukin-2 in rnalignant lymphoma. chronic lymphocytic ieukemia, and selected solid tumon. J. Clin. Oncol. 7: 75-80. 1 989.

Alvarez-Diez TM. Re- targeted Turnour I maging wi th S treptavidin 1 mmunoconjugates of Monoclonal Anti body CC49 and 1 1 '1 n-DTPA-biocytin. Master's Thesis, University of Toronto, Toronto, Cmada, 1995.

Alvarez-Diez TM. Poli hronis J. and Reill y RM. Pretargeted tumor imaging w i th streptavidin irnmunoconjugates of monoclonal antibody CC49 and 1 1 11 n-DTPA-biotin. Nucl. Med. Biol. 33: 459-66. 19%.

Andres RY, Blattmann P. Pfeiffer G, et ai. Radionuclides for therapy: A review. Schubiger PA and Haler PH (Eds.) In: Radionuclides for Therapy. Basel: Hoffman-LaRoche, 1986.

Arend WP and SilverbIatt F. Serum disappearance and catabolism of homologous immunoglobulin fragments in rats. Clin. Erp. Immunol. 72: 502-13. 1975.

Axworthy DB. Fritzberg AR. Hyalrides P, et al. Durable complete regressions of breast. lung and colon tumor xenografts with a single dose of pretargeted Y-% in a mouse model. J. Nucl. Med. 36: 2 UP, 1995.

Badger CC. Krohn KA. Peterson AV, et ai. Experimental radiotherapy of murine lymphoma wi th l3I1 labeled anti-Thy 1.1 monoclonal antibody. Cancer Res. 45: 153644, 1985.

Badger CC. Krohn KA, Shulman H, et al. Experimental radioimmunotherapy of murine lymphoma with 13'1-labeled anti-T-ce11 antibodies. Cancer Res. 46: 6273-8, 1986.

Baum RP, Niesen A, Hertel A. et ai. Activating anti-idiotypic human anti-mouse antibodies for immunotherapy of ovarian cancer. Cancer 73: 1 12 1-5, 1994.

Bayer EA, Ben-Hur H, Gitlin G, et al. An improved method for the single-step purification of streptavidin. J. Biochem. Biophp. Methods 13: 103- 13. 1986.

Begent RHJ, Bagshawe KD, Pedley RB, et al. Use of second antibody in radioimmunotherapy. Nafl. Cancer In«. Monogr. 3 : 59-6 1. 1987.

Behr TM, Sharkey RM, Juweid ME, et al. Reduction of renal uptake of radiolabeled monoclonal antibody fragments by cationic amino acids and their derivatives. Cancer Res. 55: 3825-34, 1995.

Beierwaltes WH. Radioiodine-labelled compounds previously or currently used for tumour locaiization. In: Proceedings of An Advisory Group Meeting on Tumour Localization with Rudiouciive dgenn. Panel Proceedings Section. Viema, Austria: International Atomic Energy Agency, 1974.

Bernstein ID, Eaiy JF, Badger CC. High dose radiolabeled antibody therapy of lymphoma Cancer Res. 50: 10 17s-2 1s. 1990.

Blurnenthal RD, Sharkey RM, Snyder DJ, et al. Reduction of radioantibody-induced myelotoxicity in hamsters by recombinant interieukin- 1. Cancer Res. 48: 5403-6, 1988.

Blumenthal RI>, Sharkey RM, Snyder D, et al. Reduction by anti-antibody administration of radiotoxici ty associated wi th 13 1 1-labeled anti body to carci noem bryonic antigen in cancer radioimmunotherapy . J. Nari. Cancer Insl. 8 1 : 1949, 1989.

Blumenthal RD, Kashi R. Stephens R, et al. Improved radioimmunotherapy of colorectal cancer senografts using antibody mixtures against carcinoembryonic antigen and colon-specific antigen-p. Cancer Immunof- Zmrnunother. 32: 303- 10, 199 1.

Brown 1. Astatine-211: Its possible applications in cancer therapy. Inl. J. Rad. Appl. Instriitn. 37: 789-98, 1986.

Brown JM and Lemmon W. Potentiation by the hypoxic cytotoxjn SR 4233 of ce11 hlling produced bp fractionated irradiation of mouse tumon. Cancer Res. 50: 77459. 1990.

Bruland OS. Cancer therapy with radiolabeled antibodies. An ovewiew. Acta Ortcol. 34: 1085- 94, 1995.

Buchegger F, Pèlegrin A, Delaloye B, et al. Mine- 13 1 -1abeled mA b F(ab'), fragments are more efficient and less toxic than intact anti-CEA antibodies in radioimmunothenpy of large colon carcinoma grafted in nude mice. J. Nzicl. Med. 3 1 : 103544, 1990.

Buchsbaurn DJ, Tan Haken RK, Heidom DB, et al. A cornparison of l 3 '1-labeled monoclonal antibody 17- 1 A treatment to estemal bearn irradiation on the growth of LS 174T human colon cancer senografts. Irtt. J. Rad. Oncol. Biol. Plys. 18: 1033-41, 1990.

Buchsbaum DJ, Wahl RL, Glenn SD, et al. Improved delivery of radiolabeled anti-BI monoclonal antibody to Raji 1 ymphoma senografts by predosing with unlabeled anti-B 1 monoclonal antibody. Cancer Res. 53: 637-4 1, 1992.

Buchsbaum DJ, Lawrence TS, Roberson PL, et al. Companson of I3lI- and goY-labeled monoclonal antibody 17- 1A for treatment of human colon cancer senogdts . Int. J. Radia!. Ortcd. Biol. Ph-. 25: 629-38, 1993.

Buchsbaum DJ. Esperimental radioimmunotherapy and methods to increase therapeu tic efficacv. Goldenberg DM (Ed.) In: Cancer 77ierapu W h Radiolabeled Anribodies. USA: CRC press, i995a.

Buchsbaum DJ. Espenmental approaches to increase radiolabeled antibody localization in tumors.Caricer Res. 55: 5729s-32s, 1995b.

Buras RR, Beatty BG, Williams LE, et al. Radioimrnunotherapy of hurnan colon cancer in nude mice. Ardt. Swg. 125: 66Q-4,

Buras RR, Wong JY C, Kuhn JA, et al. Cornparison of radioimrnunotherapy and estemal beam radiotherapy in colon cancer. I m . J . Rad. Oncol. Biol. P h y . 18: 103341, 1990b.

ChinoI M, Paganelli G, Sudati F, et al. Biodistribution in tumor-bearing mice of two 90Y- labelled biotins using three-s tep tumour pre-targeting. Nucl. Med. ~or%nitri. 18: 176-82, 1997.

Co bb LM and Humm JL. Radioimrnunotherapy of malignancy using antibody-targeted radionuclides. Br. J. Cancer 54: 863-70, 1986.

Colcher D, H o m Hand P. Nuti M, et ai. A spectrum of monoclonal antibodies reactive with human mammary tumor cells. Proc. N a d Acad. Sci. USA 78: 3 199-203, 198 1.

Colcher D, Esteban J, Carrasquillo JA, et al. Complementation of intracavi tary and intmvenous administration of a monoclonal antibody (B72.3) in patients with carcinoma Cancer Res. 47: 4218-34, 1987.

Colcher D, Minelli MF, RoseIli M, et al. Radioimmunolocalization of human carcinoma xenografts wi th 872.3 second generation monoclonal an ti bodies. Cancer Res. 48: 4597-603, 1988.

Crernonesi M. Chinol M. Grana C, et ai. Dosirnetric evaluations in glioma patients treated wi th yttrium-90 biotin according to a three-step pretargeting approacb J . NU&. Med. 38: 724P, 1997.

DeNardo SJ, DeNardo GL, O'Grady LF, et al. Pilot studies of radioirnmunotherapy of B cet1 1 ymphoma and leukemia using 1- 13 1-LY M- 1 monoclonal antibody. Antib. f~nnmunoconjr~g. R o d i ~ p h a m . 1 : 17-33, 1988.

DeNardo GL, DeNardo SJ, O'Grady SJ, et al. Fractionated radioirnmunotherapy of B-ce11 malignancies with 13'1-Lym- 1. Cancer Res. 50: 10 14s-6s, 1990.

DeNardo GL, DeNardo SJ, Lambom KR, et al. Enhancement of tumor uptake of monoclonal anti body in nude mice a i th PEG-IL-?. Antib. Immuttoconjug. Radiopham. 4: 859-70, 199 1.

DeNardo GL, Maddocb S W, Sgouros G. et al. Irnmunoadsorption: An enhancement strategy for radioimmunotherapy. J. Nucl. Med. 34: 1020-7, L 993.

DeNardo GL, Kroger LA, DeNardo SJ, et ai. Comparative tosicity studies of yttrium-90 MX- DTPA and 3-IT-BAD conjugated monoclonal antibody (BrE-3). Carzcer73: 10 12-27, 1994.

DeNardo GL, Kroger LA, Mirick GR, et al. Analysis of antigiobulin (HAMA) response in a group of patients wi th B-lymphocyte malignancies treated with 1311-Lym- 1. Im. J. BioL Markers 10: 67-74, 1995.

De Santes K. Slarnon D, Anderson SK, et ai. Radiolabeled antibody targeting of the HER- Ytleu oncoprotein. Cancer Res. 52: 19 16-33, 1993.

Deshpande SV, DeNardo SJ, Kukis DL, et al. Y ttrium-90-labeled monoclonal antibody for thenpy: Labeling by a new macrocyclic bifunctionai chelating agent J. Nuci. Med. 3 1: 473-9, 1990.

Dillman RO. Beauregard JC, Halpern SE, e t al. Tosicities and side effects associated with intravenous infusions of murine monoclonal antibodies. J. Biol. Response Mod. 573-84, 1986.

Dittmer J, Di ttmer A, Della Bmna R, et al. A native, affinity based protein blot for the anal ysis of streptavidin heterogeneiiy: Consequences for the specifici ty of streptavidin mediated binding assays. Ekctrophoresir 10: 763-5, 1989.

Divgi CR, Scott AM, McDermott H, et al. Clinical cornparison of radiolocalization of two monoclonal antibodies (mA bs) against TAG-72 antigen. Nucl. Med. Biol. 2 1 : 9- 15, 1994.

Divgi CR, Scott AM, Dantis L, et al. Phase 1 radioirnmunotherapy trial with M i n e - 13 1-CC49 in metastatic colon carcinoma. J. Nucl. Med. 36: 586-92, 1995.

D o m FU, Abdel-Nabi H, Baker IM, et al. Detection of pnmary coloreclal cancer with indiurn- L L L monoclonal anùbody B72.3. Arch. Surg. 125: 1601-5. 1990.

Domtngo RJ and Reill y RM. Re-tarpt ing with strepfavidin-CC49 monoclonal antibody and l 'In- and gOY-DOTA-biotin. J. Nucl. Med. 38: 188P, 1997.

Durbin PW. Metabolic charactenstics wi thin a c h e m i d family. Heaiih P h y . 2: 225-38, 1960.

Dvorak HF, Nagy JA, and Dvorak AM. Structure of solid tumors and their vascuiature: Implications for therapy with monoclonal antibodies. Cancer Ceils 3: 77-85, 1991.

Eisen HN and Keston AS. The immunologie reactivity of bovine serum albumin labeled with trace-amounts of radioactive iodine (1- 13 1). J. Immrinol. 63: 71-87, 1950.

Ellis F. The relationship of biological effect to dose-time fracfionation factors in radiotherapy. C u m f Topics in Radiation Research 4: 359-7, 1 %8.

Esteban JM, Hyams DM. Beatty BG, et al. Radioimrnunotherapy of human colon carcinomatosis xenograft wi th 90Y -2CE025 monoclonal anti bod y: Tosici ty and tumor phenotype studies. Cancer Res. 50: 989s-Es, 1990.

Ettinger DS, D r a p n LH, Klein J, et ai. Isotopic immunoglobulin in an integrated mu1 tirnodal treatment program for a primary liver cancer: A case report. Cancer Trent. Rep. 63: 13 1-4, 1979.

Ettinger DS, Order SE, W h a n m MD, e t al. Phase 1-11 study of isotopic imrnunoglobulin thcrap- for primary liver cancer. Cancer Treat. Rep. 66: 389-97, 1987.

Fowier JF. Radiobiological aspects of low dose rates in radioimmunotherapy. Irzr. J. Rodiar. Oircol. Biol. P hvs. 18: 126 1-7, 1990.

Fraker JP and Speck JC. Protein and ce11 membrane iodinations wi th a sparingly soluble chloramide 1, 3, 4, 6,-tetrachloro-3a,6a,dipheny1 glicourile. Biochern. B i o p h ~ s . Rrs. Cornrn rin. 80: 849-53, 1978.

Fujimori K, Covell, DC, Fietcher JE, et al. Modeling anal ysis of the global and rnicroscopic distri bution of I gG, F(z~b')~, and Fab in turnors. Cancer Res. 49: 5656-63, 1989.

Gailicchio VS, Hulette BC, Messino MJ, et al. Effect of various interleukins (IL- 1, IL-2, and IL-3) on the in vitro radioprotection of bone rnarrow progenitors (CN-GM and CRI-MEG). J. Biol. Response. Mod. 8: 479-87, 1989.

Gallinger S, Reiily RM, Kirsh JC, et al. Comparative dual label study of first and second generation an ti tumor-associ ated gl ycoprotei n-72 monoclonal anti bodies in colorec ta1 cancer patients. Cancer Res. 53: 271-8, 1993.

Glassy MC, Handley HH, Surh CS, et al. Geneticdly stable human hybridomas secreting tumor reactive human monoclonal IgM. Cancer lnvest. 447-55, 1987.

Goldenberg DM. Imagine and therapy of gastrointestinai cancers with radiolabeled antibodies. Am. J. Gastroenterol. 86: 1392- 1403, 1991.

Goldenberg DM. New developments in monoclonal antibodies for cancer detection and therapy. CA Cancer J. Clin. 44: 43-64, 1994.

Goodwin, DA. Pharmacokinetics and anti bodies. J. Nucl. Med. 28: 1358-62, 1987.

Green NM. The use of [MC] biotin for lunetic studies and for assay. Biochem. J. 89: 585-91, 1 963.

Green NM. A vidin. Adv. pro^. Chem. 29: 85- 133, 1975.

Greiner JW, Horand Hand P, Noguchi P, e t al. Enhanced expression of surface turnor- associated antigens on human breast and colon tumor cells after recombinant human leukocyte a-in terferon treatmen t. Cancer Res. 44: 3208- 14, 1984.

Greiner JW, Guadagni F, H o m Hand P. et al. Augmentation of tumor antigen expression by recombinant human interferons: Enhanced targeting of monoclonal anti bodies to carcinomas. Goldenberg DM and McGuire WL (Eds.) 1 n: Cancer Imaging with Radiufabeled Antibodies Boston: Martinus Nij hoff, 1990.

Greiner JW, Dansky Ullmann C, Nieroda C, et al. 1 mproved radioimmunotherapeutic efficacy of an anticarcinorna monoclonal antibody (1311-CC49) when given in combination with y- interferon. Cancer Res. 53: 600-8, 1993.

Greiner JW, Guadagni F. RoseIli M, et al. Improved experimenial radioimmunotherapy of colon xenografts by corn bining l 3 ILCC49 and interferon-y. Dis. Colon Rectum 37: S 100-5. 1994.

Fiagan PL, Halpern SE, Dillman RO, et al. Tumor size: Effect on monoclonal antibody uptake in tumor nodules. J . Nucl. Med. 37: 422-7, 1986.

Hall EJ. Radiobiologyfir the radio logis^. Philadelphia: J.B. Lippincott, 1988.

Henk JM and Smith CW. Radiotherapy and hyperbaric oxygen in head and neck cancer. Lancez : 104-5, 1977.

Henry R, Begent J. and Pedley RB. Monoclonal anti body administration. Current clinical pharrnacokinetic status and future trends. Clin. Pham. 3: 85-9, 199Z

Hericourt J and Richet C. De la serotherapie dans le traitement du cancer. C. R. Hebd. Seances Acad. Sci.. 121: 567-9, 1895.

Himmel wei t B. The Collec~ed Papers of Paul Ehrlich. London, England: Pergarnon Press, 1957.

Hnatowich DJ, Griffin TW, Kosciuckzyc C. et al. Pharmacokinetics of an I l 1-Indium-labelleci monoclonal anti body in cancer patients. J. Nucl. Med. 26: 49-58, 1985a

Hnatowich DJ, Virzi F, and Doherty PW. DTPA-coupled antibodies labeled with yttrium-90. J. Nucl. Med. 36: 503-9, 1985b.

Hnatowich DJ and McGann J. DTPA-coupled proteins: Procedures and precau tions. Int. J. Radiai. Appl. Instrwn. 14: 563-8, 1987.

Hnatowich DJ, Fritz B, Virzi F, et ai. Improved tumor locaiization with (strept)avidin and labeled biotin as a substitute for antibody. Nucl. Med. Biol. 70: 189-95. 1993.

Hikkel M, Knoop C, Schlenger K, et al. Intratumoral pOz preâicts survival in advanced cancer of the uterine cemix. Radiother. Oncol. 26: 45-50, 1993.

Humm J. Dosimetric aspects of ndiolabeled antibodies for tumor thenpy. J. Nucl. Med. 37: 1490-7, 1986.

Johnson VG, Schlom J, Paterson AJ, et al. Analysis of a human tumor-associated glycoprotein (TAG-72) identified by human monoclonal antibody 872.3. Cancer Res. 46: 850-7, 1986.

Juweid M, Neumann R, Paik C, et al. Micropharmacology of monoclonal antibodies in solid tumors: Direct experimental evidence for a binding site bamer. Cancer Res. 52: 5144-53, 1993.

Jurcic JG, Scheinberg DA, and Houghton AN. Monoclonal anti body thenpy of cancer. Pinedo HM, Longo DL, and Chabner BA (Eds.) In: Cancer Chemoîherapy arid Biologicol Response Modifirs. USA: Elsevier Science, 1994.

Kairemo KJA. Radioimmunotherapy o l solid cancers. A review. Acta Oncol. 35: 343-55, 1996.

Kalofonos HP, Ruskowski M. Siebecker A, et ai. Imaging of turnor in patients with indium- 1 1 1 -1abeIed biotin and streptavidin-conjugated antibodies: Preliminary communication. J. Nzicl. Med. 3 1 : 179 1-6, 1990.

Kantor J, Tran R. Greiner JW, et al. Modulation of carcinoembryonic antigen messenger RNA levels in human colon carcinoma cells by recombinant human y i n terferon. Caticer Res. 49: 365 1-5, 1989.

Knox SJ, Sutherland W, Goris ML. Correlation of tumor sensitivity to low-dose-rate irradiation wi th GYM phase block and other ndiobiological parameters. Radiat. Res. 135: 14- 31, 1993.

Knox SJ, Goris ML, Trisler KD, et al. goY-anti-CD?O monoclonal antibody therapy for rccurrent B cet 1 1 ymphoma. lut. J. Radial. Oncol. Biol. Phys. 32: 2 15, 1995a.

Knox SJ. Overview of studies on experi men ta1 radiotherapy. Cancer Res. 55: 5832s-6s, 199Sb.

Kobayashi H, Yoo TM, Kim IS, et al. L-lysine effectively blocks renal uptake of 12% or 99mTc-labeled anti-Tac disulphide-stabilized Fv fragment. Cancer Res. 56: 3788-95, 1996.

Kohler G and Milstein C. Continuous cultures of fused cells secreting antibody of predefined speci fici ty. k h w e 356: 495-7, 1975.

Korpela JK, Kuiomaa MS, Elo HA, et al. Biotin binding proteins in eggs of oviparous verte brates. Fxperettria 37: 1065-6, 1 98 1.

Kozbor D and Roder JC. Requirements for the establishment of high-titered human monoclonal anti bodies against tetanus toxoid using the Epstein-Barr virus technique. J. Irnrnunol. 127: 1375-80, 198 1.

Kuijpers WHA, Bos ES, Kaspersen FM, e t al. Specific recognition of antibody- oligonucleotide conjugates by radiolabeled antisense oligonucleotides: A novel approach for two step radioirnmunotherapy of cancer. Bioconj. Biochem. 4: 94- 102, 1993.

Kuzel TM and Duda RB. The current status of radiolabeled monoclonal antibodies for the diagnosis and treatment of human malignancies. Comprehensive Ther. 18: 16-10, 199Z

Lane DM, Eagle KF, Begent W. et al. Radioimmunotherapy of metastatic colorecial turnors wi th iodine- 13 1 -1abelled anti body to carcinoembryonic antigen: Phase 1/11 study with comparative biodistn bution of intact and F(abl), anti bodies. Br. J. Cancer 70: 52 1-5, 1994.

Langmuir VK and Mendonca HL. The combined use of l 3 11-labeled antibody and the hyposic cytotosin SR 4233 in vitro and in vivo. Radiat. Res. 133: 35 1-8, 1993.

Larson SM, Carnsquillo JA, Colcher DC, e t al. Estimates or radiation absorbed dose for intrapentoneally administered iodine- 13 1 radiolabeled B72.3 monoclonal antibody in patients wi th peritoned carcinomatosis. J. Micl. Med. 37: 166 1-67, 199 la.

Larson SM. Choosing the right radionuclide and antibody f o r inurtperitoneal radioirnmunotherapy. J. Natl. Caticer Itzsi. 83: 1602-4, 199 1 b.

LeBerthon B, Khawli LA, Alaudinn M, et al. Enhanced tumor uptake of macrornolecules induced by a novel vasoactive interleukin 3 immunoconjugate. Caricer Res. 5 1 : 2694-8. 199 1.

Le Doussal JM, Martin M, Gautherot E, et al. In vitro and in vivo targeting of radiolabeled monovalent and bivalent haptens wi th dual specifici ty monoclonal an ti body conjugates: Enhanced bivalent hapten dfi ni ty for cell-bound antibody conjugate. J. Nitcl. Med. 30: 1358- 66, 1989.

Lee Y C, Washbum LC, Sun TTH, et al. Radioimmunotherapy of human colorectal carcinoma senografts using -1abeled monoclonal anti body CO 17- 1 A prepared by two bi funciional chelate techniques. Cancer Res. 50: 4546-5 1 , 1990.

Leichner PK, Akabani G, Colcher D, e t ai. Patient-specific dosimetry of indium- 1 1 1- and yttrium-Wlabeled monoclonal antibody CC49. J. N~icl. Med. 38: 5 12- 16, 1997.

Leunig M. Goetz AE, Dellian M, et al. Interstitial fluid pressure in solid turnors following local hyperthermia: Possible correlation with therapeutic response. Caticer Res. 52: 4â1-90, lm.

Lybeert ML, Meerwaldt JH, Deneve W, e t al. Long-term results of low dose total body irradiation for advanced non-Hodghn's lymphoma. 1tz1. J . Radial. Oticol. Biol. Pllys. 13: 1 167-72, 1987.

Mach J-P, Buchbegger F, Bischof-Delaloye A, et al. Progress in diagnostic immuno- scintigraphy and first approach to radioimmunotherapy of colon carcinoma. Srivas tava S (Ed. ) In: Radiolabeled Momclotial Antibodies for Irnagirig and Therapv. New York: Plenum Publishing Corp., 1988.

Macklis RM, Beresford BA, Palayoor S, et al. Ce11 cycle alteraûons, apoptosis, and response to low-dose-rate radioimmunotherapy in 1 y mphoma ce1 1s. Inr. J . Radial. Oncol. Biol. Ph-. 37: 643-50, 1993.

Marshall D, Pedley RB, Boden JA, et al. Clearance of circulating radioantibodies using streptavidin o r second antibodies in a xenograft mode[. Br. J. Cancer 69: 503-7, 1994.

Mausner LF and Snvastava SC. Selection of radionuclides for radioimmunotherapy. Med. P h p 20: 503-9, 1993.

Meredith RF, Khazaeli MB, Plott WE, et al. Phase I trial of iodine- 13 1-chimeric B72.3 (Hurnan IgG4) in metastatic colorectal cancer. J. N d . Med. 33: -3-9, 1-

Meredith RF, Khazael MB, Liu T, et al. Dose fractionation of radiolabeled antibodies in patients with metastatic colon cancer. J. Nucl. Med. 33: 1648-53, 1992b.

Milenic DE, Yokota T, Filpula DR, et al. Construction, binding properties, metabolism, and tumor targeting of a single-chah Fv denved from the pancarcinoma monoclonal antibody CC49. Cartcer Res. 5 1 : 6363-7 1, 199 1.

Mock DM and Dubois DB. A spectral, solid-phase assay for biotin in physiological fluids that correlates with elcpected biotin status. Anal. Biochem. 153: 372-8, 1%.

iMoi MK, DeNardo SJ, and Meares CF. Stable bifunctional chelates of metals used in radtothenpy . Cartcer Res. 50: 789s-93s, 1990.

Mouider JE and Rockwell S. Hypoxic fractions of solid tumors: Experimental techniques, methods of analysis and a survey of existing data. Int. J . Radiat. Oacoi. Biol. Php. 10: 695- 712, 1984.

Msirikale JS, Klein JL, Shroeder J, et al. Radiation enhancement of radiolabelled antibody deposition in tumors. Inr. 3. Radiat. O~rcof. Biol. Phys. 13: 1û39-44, 1987.

Muraro R, Kuroh M, Wunderlich D, et al. Generation and characterization of B73.3 second gnen t ion monoclonal antibodies reactive with the tumor-associated glycoprotein 72 antigcn. Carrcer R a . 48: 4588-96, 1988.

Murray JL, Macey DJ, Leela PK, et al. Phase I I radioimrnunotherapy trial with 1311-CC49 in colorectal cancer. Cancer 73: 1057-66, 1994.

Murray JL, Macey DJ, Grant W. et al. Enhanced TAG-72 expression, and turnor uptake of radiolabeled monoclonal antibody CC49 in metastatic breast cancer patients following alpha interferon treatrnen t. Ca~zcer Res. 55: 3858-65, 1995.

Namh Y, Carrasquillo JA, Reynolds JC, et al. Differentiai cellular catabolism of l Wt, 90Y and 1251 radiolabeled T 10 1 anti-CDS monoclonal antibody. Nuci. Med. Bioi. 17: 10 1-7, 1990.

Neta R, Sztein MB, Oppenheim JJ, et al. The in vivo effects of interleukin 1: Bone marrow cdls are induced to cycle after administration of interleukin 1. J. Irnmunol. 139: 186 1-6, 1987.

Ngai WM, Reilly RM. Polihronis J, et al. In vitro and in vivo evaluation of streptavidin immunoconjugates of the second generation TAG-72 monoclonal antibody CC49. Nlicl. Med. Biol. 22: 77-86, 1995.

Nias NHW. An itmoduction to Radiobiology. Chichester, England: Wiley and Sons, 1990.

Nikula TK, Bocchia M. Curcio UI, et al. Impact o l high tyrosine fraction in complementarity determining regions: Measured and predicted effects o f radioiodination on IgG immunoreactivity. Mol. Immurrof. 32: 865-72, 1995.

Noungat C, Badger CC, Bernstein ID, et ai. Treatment of lymphoma with radiolabeled antibody: Elimination of tumor cells lacking target antigen. J . Natl. Cancer I m . 82: 47-50, 1990.

OtDonoghue JA. The impact of tumor ce11 prolifention in radioimmunotherapy. Cancer 73: 974-80, 1 994.

O'Donoghue JA. Optimal therapeutic s trategies for ndioimrnunotherapy. Recent Resirlfs itt Cancer Res. 14 1 : 79-99, 19%.

Okunieff P, Hoeckel M, Dunphy EP, et al. Oxygen tension distributions are sufficient to explain the local response of human breast turnors treated with radiation dose alone. Int. J . Radial. Otzcol. Biol. Phys. 26: 63 1-6, 1993.

Paganelli G, Pervez S, Siccardi AG, e t al. Intraperitoneal radio-locdi-ration of tumors pre- targeted by biotinylated monoclonal antibodies. hi. J. Cancer 45: 1 184-9, 1990a-

Paganelli G. Magnani P, Zi to F, e t al. Antibody guided tumor detection in CEA positive patients using the avidin-biotin system. J. Nucl. Med. 3 1 : 735, 1990b.

Paganelli G, Malcovati M, and Fazio F. Monoclonal antibody pretargetting techniques for tumor localization: the avidin-biotin system. Nucf. Med. Cornm. 12: ? 1 1-34, 199 la

Paganelli G, Magnani P, Zito F, et al. Three-step monoclonal antibody tumor targeting in carci noem bryoni c an ti gen-posi tive patients. Cancer Res. 5 1 : 5960-6, 1 99 1 b.

Paganelli G, Belloni C. Magnani P, et ai. Two-step targetting in ovarian cancer patients using biotinylated monoclonal antibodies and radioactive streptavidin. Eiïr. J. N i d Med. 19: 322-9. 1992.

Paganelli G, Magnani P, Meares C, et al. Antibody guided therapy of CEA positive tumors using biotinylated monoclonal antibodies, avidin and 9OY -DOTA-biotin: Initial evaluation. J. Niicl. Med. 34: 94P. 1993.

Parker BA, Vassos AB, Halpern SE, et al. Radioirnrnunotherapy of B-ce11 lymphoma with 9oY -conjugated antiidiotype monoclonal antibody. Cancer Res. 50: 1011s-8s. 1990.

Perkins AC and Pirnm MV. Administration of antibodies to patients. Perkins AC and Rmm MV ( Eds.) 1 n: Immiinoscintigraphy Practical Aspects and Clinical Applications. New York: Wiley-Liss, Inc., 199 1.

Pimm MV, Perkins AC, Annitage NC, et al. The characteristics of blood-borne radiolabels and the effcct of anti-mouse IgG antibodies on locaiization of radiolabeled monoclonal antibodies in cancer patients. J. Nlrcl. Med. 26: 10 1 1-23, 1985.

Pimm MV. Preventing rend uptake of l 1 'In from labelled monoclona1 antibody fragments. Niici. Med. Comrnritt. 16: 7 10- 1 1, 1995.

Press OW, Eary JF, Badger CC, et al. Treatment of refractory non-Hodgkin's lymphoma with radiolabeled MB- 1 (anti -CD37 anti body. 3. Clin. Oncol. 7: 1027-38, 1989.

Ress OW, Eary JF. Appelbaum FR, et al. Radiolabeled-antibody therapy of B-ce11 lymphoma wi th autologous bone rnarrow support. N. Eng. J . Med. 329: 12 19-24, 1993.

Press OW, Shan D, Howell-Clark J, et al. Comparative metabolism and retention of iodine- 11-5, yttrium-90, and indium- 1 I I ndioimmunoconjugates by cancer cells. Cancer Res. 56: 2123-9, 19%.

Rao DV and Howell RW. Time-dose fractionation in radioimmunotherapy: Implications for selecting radionuclides. J. Nucl. Med. 34: 180 1 - 10, 1993.

Reisfeld RA and Cheresh DA. Human iumor antigens. Adv. Irnrnunol. 40: 323-77, 1987.

Rosebrough SF, Iadicicco RA, and Englehart K. Quantitation, pharmacokinetics and anal ysis of biotin, avidin and streptavidin in vivo. J. Nzîcl. Med. 32: 1098, 1991.

Rosebrough SF. Pharmacokinetics and biodistri bution of radiolabeled avidin, streptavidin and biotin. Nzd . Med. Biol. 30: 663-8. 1993.

Roselli M, Schlom J, Gamow OA, et al. Comparative biodistributions of yttrium- and indium- labeled monoclonal antibody B73.3 in athymic mice bearing human colon carcinoma senograîts. J. Nucl. Med. 30: 672-82, 1989.

Rosenblum MG, Lnmki LM, Murdy JL, et al. Interferon-induced changes in phmacokinetics and tumor uptake of 1 h-labeled a&melanorna antibody %.5 in melanoma-ptients. J . Nizrl. Cnrlcer h r . 80: 1 60-5, 1 988.

Rusckowsh M. Fogarasi M. Benjamin F, et al. Effect of endogenous biotin o n the applications of streptavidin and biotin in mice. Nt~cl. Med. Biol. 24: 263-8, 1997.

Saga T, Weinstein JN, Jeong JM, et al. Two-step targeting of esperimental lung metastases wi th biotinylated anti body and radiolabeled streptavidin. Carrer Res. 54: 2 160-5, 1994.

Sands H, Jones PL, Shah SA, et ai. Correlation of tumor blood flow with monoclonal antibody uptalte by human Clouser and rend ce11 senografts. Carzcer Res. 48: 188-93, 1988.

Schechter B, Silberrnan R, Amon R, et al. Tissue distribution of avidin and streptavidin injected to mice: Effect of avidin carbohydnte, streptavidin truncation and exogenous biotin. E w . J. Biochem. 189: 327-3 1 , 1990.

Schlom J, Molinolo A, Simpson JF, et al. Advantage of dose fractionation in monoclonal anti body-targeted mdioimmunothenpy. J. Nail. Caiicer Insi. 80: 763-9, 1990.

Schlom J, Eggensperger D, Colcher D, et al. Therapeutic advantage of high-afiïnity anticarcinorna radioimrnunoconjugates. Cancer Res. 53: 1067-73, 1992.

Schubiger PA and Smith A. Optimising the radioimmunotherapy of malignant disease: the broadening choice of canier and effector rnoieties. Phurm. Acta Helveticae 34: 103- L7, 1995.

Senekowitsch R, Günther R, Mollenstadt S, e t al. Curative radioimmunotherapy of human mamrnary carci noma xenografts. J. Nucl. Med. 30: 53 1 -7, 1989.

Sharkey RM, Kaitovich FA, Shih LB, e t ai. Radioimrnunotherapy of human colonic cancer xenografts wi t h 9oY -1abeled monoclonal anti bodies to carci n o m bryonic antigen. Cancer Res. 48: 3370-5, 1988.

Shockley TR, Lin K, Nagy .iA, et al. Spatial distribution of tumor-specific monoclonal antibodies in human melanoma senogd t s . Cancer Res. 51: 367-76, 19921.

Shockely TR, Lin K, Sung C, et al. A quantitative analysis of tumor specific monoclonal anti body uptake by human melanoma xenografts: Effects of anti body immunological properties and tumor antigen expression levels. Cancer Res. 53: 357-66, 1992b.

Shrivastava S, Schlom J, Raubitschek A, et al. Studies concerning the effect of estemal irradiation on localization of radiolabeled monoclonal antibody B72.3 to human colon carci noma xenografts. Int. J. Radial. Oncol. BioL Phvs. 1 6: 73 1 -9, 1 989.

Sinitsyn VV, Mamontova AG, Ye Y, et al. Rapid b l d clearance of biotinylated IgG alter infusion of avidin. J. Nrrcl. Med. 30: 66-9, 1989.

Sneed PK and Phillips TL. Combining hypotherrnia and radiation: How beneficial?. O~cobgy 5: 99-108, 1991.

Stapely EO, Mata MJ, Miller [M. et al. Antibiotic MSD-235-1. Production by Streptomyces avidini i and Streptomyces lavendulae. Anirirnicro b. Ag. Chernother. 3: 20-7, 1 %3.

Stein R. Blumenthal R, Sharkey RM. Comparative biodistribution and radioimmunotherapy of monoclonal antibody RS7 and its F(ab')2 in nude rnice bearing human tumor xenografts. Carrcer 73: 8 16-23, 1994.

Stewart JS, Hird V, Snook D, et al. Intraperitoneal 1311- and 90Y - labelled monoclonal antibodies for ovarian cancer: Pharmacokinetics and normal tissue dosimetry. Inr. J. Cancer 3: 71 -6, 1988.

Stewart JSW, Hird V, Snook D, e t al. Intraperitoneal radioimmunotherapy for ovarian cancer: Pharmacokinetics, tosicity and efficacy of 1- 13 1 labelled monoclonal antibodies. I m . J. Radial. Oncol. Biol. Pllys. 16: 405- 13, 1989.

Stewart JSW, Hird V, Snook D, et al. Intraperitoneal yttrium-90-labeled monoclonal antibody in ovarian cancer. J. Clin. Oncol. 8: 1941 -50, 1990.

Stigbrand T, Ullén A, Sandstrom P, et al. Twenty years with monoclonal antibodies: State of the art-where do we go? Acta Olicol. 35: 39-65. 1996.

Su F-M, Gustavson LM. Axworthy DB, et al. Characterization of a new Y-90 labeled DOTA- biotin for pre-targeting. J. Nzicl. Med. 36: 154P. 1995.

Sung C, van Osdol WW, Saga T, e t al. Streptavidin distribution in metastatic tumors pretargeted wi th biotinylated monoclonal anti body: Theoretical and esperimental pharmacokinetics. Cancer Res. 54 2166-75, 1994.

Sung C and van Osdol WW. Pharmacokinetic cornparison of direct antibody targeting with pretargeting protocols based on streptavidi n-bioti n binding. J. Nucl. Med. 36: 867-76, 1995.

Thomas GD, Chappe11 MJ, Dykes PW, et al. Effect of dose, rnolecular size, affinity and protein binding on tumor uptake of anti body or ligand: a biomathematical model. Cancer Res. 49: 3290-6, 1989.

Thor A, Ohuchi N, Szpak CA. et al. Distribution of oncofetal antigen tumor-associated glycoprotein-72 defined by monoclonal antibody B72.3. Cancer Res. 46: 3 1 18-24, 1986.

Ullén A, Nilsson B. Ricklund Ahlstrom K, et al. In vivo and in vitro interactions between idiotypic and antiidiotyptc monoclonal antibodies against placentai alkaline phosphatase. J. Immrîrzol. Methods 183 : 155-65, 1 995.

Urtasun R, Band P. Chapman JD, et al. Radiation and high-dose metronidazole in supratentonal glioblastomas. N. E q . J . Med. 294: 1364-7, 1976.

van Osdol WW, Sung C, Dednck RL, et al. A distnbuted pharrnacokinetic model of two-step imaging and treatment protocols: Application to streptavidin conjugated monoclonal antibodies and radiolabeled biotin. J. Nucl. Med. 3 4 1552-64, 1993.

Vriesendorp HM, Quadri SM, Stinson RL, et al. Selection of reagents for human ndioimmunotherapy. h t . J . Radial. O~rcol. Biol. Ph-. 22: 37-45, 199 1.

Vriesendorp HM, Quadri SM, Borje S. et al. Hematologic side effects of radiolabeled i mmunoglobulin therapy. Erp. Hernatol. 24: 1 183-90, 19%.

Wagle SR. Hofman F, and Decker K. Studies on protein degradation in isolated hepatocytes and Kupffer cells. Segal HL and Doyle DJ (Eds.) In: Protein Turnover and Lysosome F~~nction. London: Academic Press, pp. 7 15-29. 1978.

Wahl RL, Parker CW, and Philpott GW. Improved radioimaging and tumor localizaiion wi th monoclonal F(ab')z. J. Nucl. Med. 24: 3 16-25, 1983.

Wang S. Quadri SM, Tang XZ, et al. Liver tosicity induccd by combined esternal beam imdiation and radioirnmunoglobulin therapy. Radial. Res. 14 1 : 294302, 1995.

Weiden PL, Wolf SB, Breitz HB, et al. Human anti-mouse antibody suppression with cyclosporin A. Cancer 73: 1093-7, 1994.

Welt S, Divgi CR, Kemeny N, et al. Phase 1/11 study of iodine 13 1 -1abeled monoclonal anti body A33 in patients wi th advanced colon cancer. J. Clh. Orzcol. 12: 156 1-7 1, 1994.

Wessels BW and Rogus RD. Radionuclide selection and model absorbed dose calculations for radiolabeled tumor associated antibodies. Med. Ph-. 1 1 : 638-42, 1984.

Wessels BW, Vasella RL, Palme DF, et al. Radiobiological companson of extemal beam irradiation and radioimmunotherapy in renal ce11 carcinoma xenografts. Inl. J. Rad. Oncol. Biol. Phys. 17: 1357-63, 1989.

Wilchek M and Bayer EA. The avidin-biotin corn plex in bioanal ytical applications. Amlyiical Bioclzerrr. 171: 1-33, 1988.

Wilder RB, Langmuir VK, Mendonca HL, et al. Local hyperthemia and S R 4233 enhance the antiturnor effects of radioimmunotherapy in nude mice with human colonic adenocarcinorna senogtafts. Cancer Res. 53: 3022-7, 19%.

Wilder RB, McGann JK, Sutherland WR, et al. The hypoxic cytotoxin SR 4233 increases the effectiveness of radioimmunotherapy in mice wi th non-Hodgkin 's 1 ymphoma s e n o g d t s . Itir. J . Rudiat. Oncol. Biol. P h p . 23: 1 19-76, 1994.

Wilder RB, DeNardo GL, and DeNardo SJ. Radioimm unotherapy: Recent resul ts and future directions. J. Clin. Ortcol. 14: 1383-1400, 1996.

Williams LE, Duda RB, Proffitt RT, et al. Tumor uptake as a function of tumor mas: A mathematicai model. J. Nucl. Med. 29: 103-9, 1988.

Woolley DW and Longsworth LG. Isolation of an antibiotin factor from egg white. J. Biol. Chem. 143,: 285-90, 1943.

Yokota T, Milenic DE, Whitlow M. et al. Rapid tumor penetration of a single-chain Fv and comparison with other immunoglobulin forms. Cartcer Res. 5 2 3401-8, 1991.

Zalutsky MR a n d B igner DD. Radioimmunotherapy w i th a-particle emitting radioimmunoconjugates. Acia O~icol. 35: 373-9, 19%.

Zhang M. Yao 2, Sakahara H, et al. Radiotherapy of intraperitoneal tumor wi th iodine- 13 1 labeled avidin. J. Nucl. Med. 38: 194P, 1997.

Zimmer A, Rosen ST, Spies SM, et al. Radioimmunotherapy of patients with cutaneous T-ce11 lymphorna using an iodine- 13 1 -1abeled monoclonal antibody: Analysis of retratment folloiving plasrnapheresis. J. Nucl. Med. 29: 174-80, 1988.

LIST OF PUBLICATIONS AND ABSTRACTS

RM Reilly, RJ Domingo, and J Sandhu. Oral delivery of antibodies: Future pharmacokinetic trends. Clinical P~tantlclcokinetics 32: 3 13-31, 1 997.

RI Domingo and Rh4 Reilly. Pre-targeting with streptavidin-CC49 monoclonal antibody and 1In- and gOY-DOTA-biotin. Journal of Nuclear Medicine 38(Suppl. 5): 188P, 1997.

(A bsmt)

APPENDICES

APPENDIX 1

Purification of CC49 Monoclonal Antibody From Mouse Ascites Fluid

Mouse ascites fluid was generously provided by Dr. J. Schlom of The National Cancer

Institute, Bethesda, Maryland. CC49 m A b was purified from ascites fluid by affinity

chromatography using a Protein G agarose col um n (Pierce, Rockford. 1 L). Followi ng

purification from ascites fluid, CC49 mAb was desalted on a PD- IO Sephadex G-15 column

( P h m a c i a , Baie dlUrfe. Québec) eluted with sodium bicarbonate bufier (50 mM, pH 7.5) and

then concentrated to 10 mgImL in a Centricon-30 concentrator (Amicon, Beverly, MA). The

punty of CC49 mAb was evaluated by size-exclusion HPLC on a Bio-Silect 750-5 column

(Bio-Rad, Richmond, CA) eluted with phosphate buffer (50 mM, pH 6.5) at I mUminute.

Protein was identified by using an in-line UV absorbance monitor (Mode1 1141, LKB,

Bromma, Sweden) set at 215 nm. The molecular weight of the purified protein ivas estimated

using a standard curve, generated from the retention times of protein standards. The protein

standards ( Bio-Rad) incl uded: tyroglobulin (670 D a ) . bovine gamma glo bulin ( 158 kDa).

oval bumin (44 D a ) , myoglobin ( 17 ma), and cyanocobdamin ( 1.3 ma), which had retention

times of approsimately 6.4.8.8, 9.5, 11.4, and 13.6 minutes, respectively. CC49 mAb purity

was also eu.luated by SDS-PAGE using a 420% Tris-glycine gel (Bio-Rad). Protein bands

rvere identified by Coomassie R-750 stain.

APPENDIX II

Preparation of Streptavidin-CC49 Monoclonal Antibody

CC49 mA b purified from mouse ascites fluid was conjugated to streptavidin (Sigma

Chemical Co., St. Louis. MO) using sulfosuccinimidyl4(N-maleimidomethy1)cyclohexme- 1 -

carboxylate (sulf+SMCC) (Pierce Chernical Co.), a protein cross-linker rnolecule (Ngai et al.,

1995; Alvarez-Diez et ai., 1996). Bnefly, disulphide bt-idges on the CC49 mAb (10 mglmL)

were reduced by incubation wi th mercaptoethy lamine (Sigma Chemical Co.) a t a

mercaptoethylarnine:CC49 mAb molar ratio of 6.000: 1 in phosphate buffered saline (PBS) (M

rnM; pH 7.3) for I hour and 30 minutes at a tempenture of 37" C. Strepiavidin ( 1 mg/mL) was

reacted with sulfo-SMCC at a sulfo-SMCCstreptavidin molar ratio of 1,440: 1 in PBS for 30

minutes at a temperature of 37" C. Reduced CC49 mAb and maleimide-derivatized streptavidin

were then purified from escess reagents on a Sephadex G-M mini-column eluted wi th PBS.

Sephades G-M mini-columns were prepared by packing glass Pasteur pipettes, plugged at the

neck with glass wool, with Sephades G-50 resin (Sigma Chemical Co.). After purification,

reduced CC49 mA b and maleimide-derivatized streptavidin were reacted together ovemight at

room temperature. The molecular weight and puri ty of the conjugate were evaluated by size-

esclusion HPLC and by SDS-PAGE, as previously described in Appendis 1.

APPENDIX III

Measurement of the Immunoreactive Fraction of Streptavidin-CC49 Monoclonal Antibody

The immunoreactive fraction of streptavidin-CC49 was determined by rneasunng the

fraction of radiolabeled conjugate bound to CNBr-activated Sepharose 4B beads (Sigma

Chemical Co.) that were previously covalent1 y linked to bovine submaxillary mucin (BSM)

(Sigma Chemical Co.), a known source of the TAG-77 antigen. Briefl y, streptavidin-CC49

was incubated wi th 1 1In-DOTA-biotin I'or 30 minutes at a temperature of 37'C and the

unbound 1 1 1 In-DOTA-biotin was removed by filtration using Centricon-30 concentrators.

Approximately 30 ng of streptavidin-CC49, labeled with 1 IlIn-DOTA-biotin, was incubated

with 125 pL of BSM-coated Sepharose beads in a 0.45 pm Micropure Filter (Amicon) for 45

minutes at room t e m p n t u r e under gentte agitation. At the end o l the incubation period, the

filter was counted for 1 "ln radioactivity in a gamma well counter (Packard Mode1 Auto-

Gamma 5650, Downer's Grove, IL) using a window of 1 9 - 2 6 0 keV to include the 173 keV

and 740 keV y-photons of l In. The fi l ter ivas then centrifuged and the fil tnte containing

unbound streptavidin-CC49 labeled with l 1 'In-DOTA-biotin was incubated cvith a new 115 PL

aliquot or BSM-coated Sepharose beads, as previously described. After the completion of the

second incubation, the f i l ter was centrifuged, the fi l trate discarded, and both fil ters containing

BSM-coated Sepharose were counted for 1 1 'In radioactivity. Based on the known amount ol'

ridioactivi ty added to the BSM-coated Sepharose beads in the first incubation, the

immunoreactive fraction of streptavidin-CC49 was measured as the total percentage of activity

remaining on the fi 1 ters.

APPENDIX IV

Prepamtion of 1 1 iIn- and 9 OY-DOTA-biotin

DOTA-biotin was generously provided by the NeoRx Corporation of Seattle,

Washington. 1 1 'In- and 9OY -DOTA-biotin were prepared by incubating DOTA-biotin with

1 1 )In chloride (Nordion, Kanata, Ont.) o r chloride (New England Nuclear, Boston, MA)

in ammonium acetate buffer (500 mM, pH 5) for 60 minutes in a water bath heated to a

temperature of 80°C. The radiochernical puri ty of 1 1 In- and -DOTA-biotin was determined

by reversed-phase chromatography using disposable C- 18 Sep-Pak cartridges (Waters,

Milford, MA). For pre-iargeting studies, the streptavidin-binding capacity of radiolabelcd

DOTA-biotin was evaluated by incubating radiolabeled DOTA-biotin with a 24-f-old molar

excess of streptavidin-CC49 for 30 minutes at a temperature of 37'C. Streptavidin-CC49

bound to radiolabeled DOTA-biotin and free ndiolabeled DOTA- biotin were sepanted by

clution through a Sephades G-50 mini-column with 0.9% saline. The eluate fractions were

measured for 1 1 l ln radioactivity in a gamma well counter (Packard Mode1 Auto-Gamma 5650)

as previously described in Appendis III . To rneasure 90Y radioactivity. the eluate fractions

werc measured for Bremstrahlung radiation in the gamma well counter using an open window

of M- 1000 keV. The streptavidin-binding capacity was measured by comparing the total

radioactivity eluted in the void volume fractions ( ' 1 1111-IgoY-DmA-biotin-strepiavidin-CC49)

to that eluted as free radiolabeled DOTA-biotin in the fractionation range eluates.

APPENDIX V

Establishment of LS174T Colon Cancer Xenografts

LS 174T human colon cancer cells were obtained [rom the Arnerican Type Cell Culture

(Rockville, MD) Company and maintained in Eagle's Minimum Essential Medium

supplemented with 10% fetal calf serum (Sigma Chemical Co.) under 37'C/5% CO2

conditions. For x e n o g d t establishment. LS 174T cells were recovered by trypsinization with

0.25% trypsin1EDTA (Sigma Chemical Co.) and resuspended to 3.0 x IO7 cellslmL in stenle

0.9% saline. Appronimately 3.0 s 106 LS 174T cells in a 100 PL volume were injected

subcutaneously in the lower left abdominal region of six- to eight -week old femaie Swiss

nuhu (athymic) mice (Charles River Laboratoi-ies, Boston. MA) and ailowed to grow for 8- 15

days pnor to the start of the biodistnbution and thenpy studies. Ali mice were acclimatized for

at least two weeks before senograft establishment and were housed in sterile c a p s fi tted wi th

fïltered lids. The mice were maintained on a diet of stenle water and rodent chow (Autoclavable

Rodent Chow 50 10, PMI Feeds, St. Louis, MO). Al1 animal studies were conducted under an

approved proiocol from the Animal Care Committee at The Toronto Hospi ta1 following CCAC

guideIines.

APPENDIX VI

Calculation of Absolute Streptavidin-CC49 and DOTA-biotin Uptake in LS 174T Xenografts

DOTA- biotin pmois DOTA- biotidg of tumour ( A) = ((% iidg of tumour) i 100) x mass of injectatel ( B )

=((B) + 900Da) x lxlO1~prnolslmol

Sîreptavidirr - CC49 Morwclo~ral hrtibody pmols streptavidin-CC491g of tumour (C) = ((% idlg + 1100) x 0.5t x mass of injectate$ ( D )

= ((D) + 215 000 DaY) s ls10L2 pmols/mol

Therefore, DOTA-biotin:streptavidin-CC49 ratio = (A) i ( C )

DOTA-biotirr (Nomlized IO Blood) pmols DOTA-biotidg of blood (E) = ((% id/@ + 1 0 ) s mass of injectates (F)

= ((F) + 900 Da) s 1 x 101 2 pmols/mol

Therefore. absoiute tumour uptake of DOTA-biotin (nomaiized to blood) (G) = (A) + (E)

Strepraviditr -CC49 Monoclorzal Atltibody (Norrnolked to Blood) pmols streptavidin-CC49lg of blood (H) = ((9% idlg) + 100) s O.Si s mass of injectatef (1)

= ((1) + 2 15 000 DaY) s lx 101 pmolslmol

Therefore, absoi ute tumour u p u e of streptavidi n-CC49 (normaiized to blood) (J ) = (C) s (H)

Therefore, DOTA-bi0tin:streptavidin-CC49 ratio (nomalized to blood) = (G) + (J)

rmass = ls106, 2slC5. 4s1Ck5, or 1.5slû-4g ' 50% of injectate is unconjugated CC49 mA b

mass = 3.5s lû-4 g Streptavidin-CC49 presumed to be made up of 1 streptavidin molecule (65 kDa) and 1 CC49

mA b ( 1 M D a ) . Therefore, total molecular weight = ? 15 kDa.

IMAGE EVALUATION

APPLIED 4 IMAGE. lnc - = 1653 East Marn Street - -A Rochester, NY 14609 USA -- -- - - Phone: 71 W482-0300 -- -- - - Fa: 7161288-5989