[CANCER RESEARCH 52, 5693-5700, October 15, 1992]

Antigen-specific Activity of Carcinoma-reactive BR64-Doxorubicin Conjugates

Evaluated in Vitro and in Human Tumor Xenograft ModelsPamela A. Trail,1 David \VMiner, Shirley J. Lasch, Arris J. Henderson, Robert S. Greenfield, Dalton King,

Mary E. Zoeckler, and Gary R. Braslawsky.

Department of Experimental Therapeutics [P. A. T., S. J. L., A. J. H.J, Department of Immunology [R. S. G., M. E. Z., (l. R. B./, and Department of Chemistry[D. W., D. K.J, Bristol-Myers Squibb Pharmaceutical Research Institute, Wallingford, Connecticut 06492

ABSTRACT

The anticarcinoma antibody BR64 was conjugated to a doxorubicinderivative, doxorubicin 13-[3-(2-pyridyldithio)propionyl]hydrazone, andthe resulting conjugates (BR64-DOX) were evaluated for activity andimmunological specificity in vitro and in human tumor xenograft models. The BR64-DOX immunoconjugates retained immunoreactivity andcytotoxicity and demonstrated antigen-specific cytotoxicity in vitro. Thepotency of BR64-DOX immunoconjugates in vitro was related to thedrug:monoclonal antibody mole ratio of the conjugates. The antitumoractivity of BR64-DOX conjugates was consistently superior to the maximal activity obtained with the parent drug, doxorubicin (DOX), inestablished human lung and human breast carcinoma xenograft models.The superior antitumor activity of BR64-DOX conjugates was reflectedboth in tumor growth inhibition and in regressions and cures of established tumors following the administration of tolerated doses of BR64-DOX. The antitumor activity of BR64-DOX conjugates was not theresult of synergism between monoclonal antibody BR64 and DOX, because mixtures consisting of monoclonal antibody and optimized DOXwere not more active than an equivalent dose of DOX administeredalone. The antitumor activity of BR64-DOX conjugates was antigenspecific; equivalent doses of nonbinding isotype-matched conjugateswere not active against established tumor xenografts.

INTRODUCTION

Successful treatment of solid tumors is a difficult therapeuticchallenge. Although a variety of chemotherapeutic drugs havebeen shown to be active, the therapeutic index of these agents is,in general, low due to nonspecific toxic effects on normal tissues. One promising means of increasing the efficacy of chemotherapy is to utilize MAbs2 directed against tumor-associ

ated antigens to selectively deliver cytotoxic agents to tumorsites. Site-directed delivery may increase the intratumor concentration of the cytotoxic agent while decreasing systemic tox-icity. Recently, the concept of MAb-directed delivery has received significant attention and MAbs have been conjugated toa variety of toxic moieties, including chemotherapeutic agents(1-9), radionuclides (10-12), and plant (13-15) and bacterial(16-18) toxins.

The use of MAb-directed delivery may significantly reducethe toxic side effects of therapy. However, because the majorityof MAbs available react to some degree with normal cells, it iscritical to develop immunoconjugates in which normal tissuereactivity will not produce unacceptable levels of toxicity. Forthis reason immunoconjugates produced with clinically known

Received 4/21/92; accepted 8/5/92.The costs of publication of this article were defrayed in part by the payment of

page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1To whom requests for reprints should be addressed, at Bristol-Myers SquibbPharmaceutical Research Institue. P.O. Box 4000, Princeton, NJ 08543-4000.

2 The abbreviations used are: MAb, monoclonal antibody; DOX, doxorubicin;DOXHZN, doxorubicin 13-[3-(2-pyridyldithio)propionyl]hydrazone; PCS, fetalcalf serum; MTD, maximum tolerated dose; TVDD, tumor volume doubling delays; q4dx3, every 4 days for three injections; q7dx2, every 7 days for two injections;q2dx5, every 2 days for five injections; IC50, 50% inhibitory concentration.

chemotherapeutic drugs offer a significant safety advantageover more potent, but less well defined, agents such as the plantand bacterial toxins. The conventional chemotherapeutic drugshave established clinical profiles of antitumor activity and thedose-limiting toxicities of the drugs have been well character

ized. In particular, anthracyclines such as DOX have beenshown to have a broad spectrum of activity against solid tumors, and the toxic effects of the drug are dose related andpredictable (19, 20). A variety of immunoconjugates have beenproduced in which anthracyclines are linked to MAbs usingseveral different conjugation strategies (1, 3, 4, 9, 21-28).

In the present study, DOXHZN (26, 27), which contains anacid-labile acyl hydrazone bond, was conjugated to an anticarcinoma MAb termed BR64. The BR64 MAb is a murine IgGlantibody which identifies a Ley-related tumor-associated anti

gen expressed on the surface of carcinoma cells of the breast,colon, lung, and ovary. The antibody is internalized followingantigen-specific binding (29). This report describes the antigen-specific activity of BR64-DOX immunoconjugates in vitro andthe antitumor activity of BR64-DOX compared to that of theoptimized parent compound, DOX, in human carcinomamodels.

MATERIALS AND METHODS

Monoclonal Antibodies. MAb BR64 (IgGl) identifies a Le>-relatedtumor-associated antigen expressed on carcinomas of the lung, colon,

breast, and ovary (29). The SN7 hybridoma line was received from B.Seon (Roswell Park Memorial Institute, Buffalo, NY). The SN7 MAbidentifies an antigen expressed on human B cells and was used in thesestudies as a nonbinding, isotype-matched, control antibody. Both theBR64 and SN7 MAbs were produced as tissue culture supernatants(Brunswick BioTecnetics, San Diego, CA).

Tumor Cell Lines. L2987 is a human lung adenocarcinoma linewhich expresses the BR64 antigen, and it was obtained from I. Hell-strom (Bristol-Myers Squibb, Seattle, WA). MCF7 is an estrogen-dependent human breast carcinoma line which expresses the BR64 antigen, and it was obtained from the American Type Culture Collection.HCT116 is a human colon carcinoma line, obtained from M. Brattain(Baylor College of Medicine, Houston, TX), which does not express theBR64 antigen.

Synthesis of MAb-DOX Immunoconjugates. DOXHZN and DOXconjugates were prepared as described previously (26, 27). Briefly, immunoconjugates were prepared by linking DOXHZN to MAbs whichhad been thiolated with n-succinimidyl 3-(pyridyIdithiol)proprionate.The thiopyridyl protecting groups were removed from the MAb byreduction with excess dithiothreitol, and the free thiol-containingMAbs were condensed with DOXHZN. Unconjugated drug was removed by dialysis and passage over SM-2 Biobeads (Bio-Rad Laboratories, Richmond, CA). Immunoconjugates were evaluated for free drugcontamination by high performance liquid chromatography and forretention of MAb binding activity as described previously (30). Immunoconjugates used in these studies contained <5% free DOX andretained >90% of the original antibody binding activity. Immuno-

5693

on May 26, 2021. © 1992 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTITUMOR ACTIVITY OF BR64-DOX CONJUGATES

conjugates are denoted by the conjugate mole ratio (moles drug/moleimmunoglobulin).

In Vitro Cytotoxicity Assays. Monolayer cultures of the varioushuman carcinoma cells were harvested using trypsin-EDTA (GIBCO,Grand Island, NY), and the cells were counted and resuspended to 1 X105/ml in RPMI 1640 containing 10% heat-inactivated fetal calf serum(RPMI-10% PCS). Cells (0.1 ml/well) were added to each well of 96-well microtiter plates and incubated overnight at 37°Cin a humidified

atmosphere of 5% CO2. Medium was removed from the plates andserial dilutions of DOX or MAb-DOX conjugates were added to the

wells. All dilutions were performed in quadruplicate. Cells were exposed to DOX or MAb-DOX conjugates for 2 h at 37°Cin a humidified

atmosphere of 5% CO2. Plates were then centrifugea (200 x g, 5 min),the drug or conjugate was removed, and the cells were washed 3 timeswith RPMI-10% FCS. The cells were cultured in RPMI-10% PCS(37°C,5% CO2) for an additional 48 h. At this time the cells werepulsed for 2 h with 1.0 ¿iCi/wellof [3H]thymidine (New England Nu

clear, Boston, MA). The cells were harvested onto glass fiber mats(Skatron Instruments, Inc., Sterling, VA) and dried, and filter-bound[3H]thymidine radioactivity was determined (0-Plate scintillation

counter; Pharmacia LKB Biotechnology, Piscataway, NJ). Inhibition of[3H]thymidine uptake was determined by comparing the mean cpm for

treated samples with the mean cpm for the untreated control.Experimental Animals. Congenitally athymic female mice of

BALB/c background (BALB/c nulnu; HaríanSprague-Dawley, Indianapolis, IN) were used in these studies. Mice were housed in Thorencaging units on sterile bedding with controlled temperature and humidity. Animals received sterile food and water ad libitum.

Human Tumor Xenograft Models. The L2987 and MCF7 humantumor lines were established as tumor xenografts in athymic mice. Alltumor lines were maintained by serial passage as described previously(31). The MCF7 tumor is an estrogen-dependent human breast tumorline. Athymic mice were implanted with 0.65-mg (65-day release rate)estradici pellets (Innovative Research of America, Toledo, OH) 1 dayprior to the implantation of MCF7 tumors. L2987 and MCF7 tumorswere measured in two perpendicular directions at weekly or biweeklyintervals, using calipers. Tumor volume was calculated according to theequation: V = I x w2/2, where V = volume (mm3), / = measurement oflongest axis (mm), and w = measurement of axis perpendicular to /.

In general, there were 8-10 mice in each control or treatment group.Data are presented as median tumor size for control or treated groups.Antitumor activity is expressed in terms of median TVDD values,where TVDD = T - C/TVDT; T - C is defined as the median time(days) for treated tumors to reach 500 mm3 in size minus the mediantime for control tumors to reach 500 mm' in size and TVDT is the time(days) for control tumors to double in volume (250-500 mm3). Partialtumor regression reflects a decrease in tumor volume to <50% of theinitial tumor volume, complete tumor regression refers to a tumorwhich for a period of time is not palpable, and cure is defined as anestablished tumor which is not palpable for a period of time of a 10TVDTs.

For animals bearing the L2987 human lung tumor the averageTVDT was 4.8 ±0.9 days and therapy was initiated when the mediantumor size was 75 mm3 (typically 14 days after tumor implantation).For animals bearing the estrogen-dependent MCF7 breast tumor theaverage TVDT was 6.4 ±2.0 days and therapy was initiated when themedian tumor size was 100 mm3 (typically 13 days after tumor implan

tation).Therapy. Treatments were administered by the i.p. or i.v. route on

various schedules, as denoted. DOX was diluted in normal saline andMAb and MAb-DOX conjugates were diluted in phosphate-buffered

saline. All therapy was administered on a mg/kg basis, as calculated foreach animal, and doses are presented as mg/kg/injection. Control animals were not treated. Doses of immunoconjugate are reported basedon the drug (equivalent DOX) and antibody content. The MTD for atreatment regimen is defined as the highest dose on a given schedulewhich resulted in <20% lethality.

RESULTS

Antigen-specific Cytotoxicity of BR64-DOX Conjugates inVitro. Several human carcinoma lines, representing varioushistological types, express the BR64 antigen (29). Among these,the L2987 lung and MCF7 breast carcinoma lines were selectedfor use in these studies. As shown in Fig. 1, A and A, BR64-DOX conjugates demonstrated antigen-specific Cytotoxicity invitro when evaluated against the L2987 and MCF7 human carcinoma lines, respectively. The BR64-DOX conjugate was approximately 15-fold more potent than the SN7-DOX conjugate(IC50 values of 2.0 and 30 MMDOX, respectively) against theL2987 line. Antigen-specific Cytotoxicity was also demonstrated by blocking of the activity of the BR64-DOX conjugatewith the addition of unconjugated BR64 antibody. As shown inFig. IA, the potency of the BR64-DOX conjugate (IC50 of 2 MM)was significantly reduced in the presence of MAb BR64 (ICSOof20 MM)but was not affected (IC50 of 2 MM)by the addition of anisotype-matched nonbinding antibody.

The BR64-DOX conjugate was approximately 25 times morepotent than the SN7-DOX conjugate (IC50 values of 0.2 and 5MMDOX, respectively) against the MCF7 breast carcinoma line(Fig. IB). The BR64-DOX conjugate was not active (IC50 of>30 MM)against the HCT116 carcinoma cell line (Fig. 1C).This cell line is sensitive to unconjugated DOX (IC50 of 0.04MM)but does not express the BR64 antigen.

To investigate the relationship between the mole ratio of drugto MAb and in vitro potency of BR64-DOX conjugates, a variety of conjugates with conjugate ratios ranging from 1 to 8were prepared. These immunoconjugates were evaluated invitro against the L2987 lung carcinoma line. The potency of theimmunoconjugates varied over a 33-fold range (IC50 values of1-33 MMDOX). As shown in Fig. 2, conjugates of higher moleratio were significantly (P < 0.05) more potent in vitro thanthose conjugates prepared at lower mole ratios.

In Vivo Antitumor Activity of DOX against Established Tumor Xenografts. The optimal route, dose, and schedule ofDOX administration were determined for established L2987and MCF7 tumor xenografts. The antitumor activity of DOXevaluated against established L2987 tumor xenografts is summarized in Table 1. Therapy was initiated approximately 14days after tumor implantation, when the tumors were 50-100mm3 in size. The MTD of DOX was dependent on the schedule

and route of administration. With i.p. administration the MTDof DOX was 4 mg/kg given q4dx3 or 8 mg/kg given q7dx2.DOX administered i.p. at tolerated doses was not active againstestablished L2987 tumors. The MTD was significantly higherwhen DOX was administered by the i.v. route. Maximal anti-tumor activity was observed when DOX was administered bythe i.v. route either at a dose of 8 mg/kg on a q4dx3 schedule orat a dose of 10 mg/kg on a q7dx2 schedule. At these doses theantitumor activity was equivalent to 4.5 ±0.7 and 4.6 ±0.7TVDDs, respectively. Regressions and cures of established tumors occurred in <5% of animals treated with tolerated dosesof DOX.

The antitumor activity of DOX tested against establishedMCF7 tumors is presented in Table 2. Maximal antitumoractivity was observed when DOX was administered i.v. at a doseof either 6 mg/kg on a q4dx3 schedule or 8 mg/kg on a q7dx2schedule. At these doses the antitumor activity was equivalentto 3.0 ±0.4 and 2.9 + 0.6 TVDDs, respectively. Regressionsand cures of established tumors were not observed. The MTD

5694

on May 26, 2021. © 1992 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTITUMOR ACTIVITY OF BR64-DOX CONJUGATES

100

0.001 0.01 0.1 1

Doxorubicm (uM)

100

O 100 -i

100

0.001 0.01 100

Doxorubicm (uM)

Fig. 1. Cytotoxicity of BR64-DOX conjugates against L2987 lung (A), MCF7breast (A), and HCT116 colon (Q carcinoma lines. Cells were exposed to DOXor DOX conjugates for 2 h at 37°C,washed, and cultured for an additional 48 hprior to the addition of [3H]thymidine. Data are presented as percentage ofinhibition (mean ±SD) of [3H]thymidine uptake, relative to control cells. A,L2987 cells treated with DOX (A), BR64-DOX (mole ratio = 6.74) (•),ornonbinding SN7-DOX (mole ratio = 5.52) (O). The effect of adding unconjugatedMAb BR64 (•)or a nonbinding isotype-matched MAb (D) to BR64-DOX conjugates is also shown. B, MCF7 cells treated with DOX (A), BR64-DOX (moleratio = 4.88) (•),or nonbinding SN7-DOX (mole ratio = 3.89 (O). C, HCT116cells treated with DOX (A) or BR64-DOX (mole ratio = 6.74) (•).

Conjugate Mole Ratio

Fig. 2. Relationship between conjugate mole ratio and in vitro potency ofBR64-DOX conjugates. Conjugates were prepared with mole ratios of 1-8 andwere tested against L2987 lung carcinoma cells. Cells were exposed to BR64-DOX conjugates for 2 h at 37"C, as described in "Materials and Methods." Dataare presented as the IC50 value (mean ±SD) of BR64-DOX conjugates.

of DOX is lower in the MCF7 tumor model than in the L2987tumor model as a result of the estrogen supplementation required for in vivo growth of the estrogen-dependent MCF7tumor. A similar decrease in the MTD of a variety of cytotoxicdrugs has been observed in estrogen-supplemented mice (datanot shown). Prolonged exposure of mice to exogenous estrogenhas been shown to produce hepatic and renal lesions (32) andmay result in a decrease in MTD by reducing the capacity todetoxify and excrete cytotoxic agents.

Antitumor Activity of BR64-DOX Immunoconjugates Evaluated against Established Human Tumor Xenografts. The anti-tumor activity of BR64-DOX conjugates was evaluated againstestablished (50-100-mm3) L2987 human lung tumor xe-

nografts and compared to that of optimized DOX and equivalent doses of nonbinding immunoconjugates. Mice were treatedi.v. with optimized DOX or i.p. with BR64-DOX. Data presented in Fig. 3A compare the antitumor activity of DOX administered at the MTD on a q4dx3 schedule with that of aBR64-DOX conjugate (mole ratio = 5.37) administered i.p. on

the same schedule. DOX (8 mg/kg) produced a significant delayin the growth of established L2987 tumors. The antitumoractivity of DOX in this experiment was equivalent to 5.0TVDDs, and tumor regressions were not observed. The BR64-DOX conjugate tested at the MTD (35 mg/kg equivalent DOX)was significantly (P < 0.01) more active than optimized DOXand produced activity equivalent to 11.9 TVDDs. At this dose63% partial regressions and 13% cures were observed. Administration of the BR64-DOX conjugate at a lower dose (25mg/kg equivalent DOX) resulted in antitumor activity equivalent to 10.6 TVDDs. At this dose 33% partial regressions and17% cures of established tumors were observed. These animalsare long term cures and the mice have remained tumor free for>1 year.

The antitumor activity of BR64-DOX conjugates was alsocompared to that of nonbinding (SN7-DOX) conjugates administered at equivalent doses and with the same schedule and

5695

on May 26, 2021. © 1992 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTITUMOR ACTIVITY OF BR64-DOX CONJUGATES

Table 1 Antitumor activity of DOX against established L29S7 human lung carcinoma xenografts

Tumor regressions(%)Schedule,

routeq4dx3.

i.v.q7dx2,

i.v.Single

injection,i.v.q4dx3,

i.p.q7dx2,

i.p.Dose

(mg/kg)1086412108161412642863TVDDToxic4.5

±0.7*3.6

±0.31.3±0.7Toxic4.6

±0.73.0±0.3ToxicToxk3.0Toxic0.4

±0.10.41.00.40.3Partial3.60000000000Complete00000000000Cure0001.12.1000000Lethality

(%)47(19)"2(55)0(46)0(20)33

(63)2(94)0(48)100(10)33

(9)22(9)32(19)0(19)0

(8)0

(9)0(8)0

(8)" Values in parentheses, numbers of mice.* Mean ±SD.

Table 2 Antitumor activity ofDOX against established MCF7 human breast carcinoma xenografts

Tumor regressions (%)

Schedule, route Dose (mg/kg) TVDD Partial Complete Cure

" Values in parentheses, numbers of mice.* Mean ±SD.

Lethality (%)

q4dx3, i.v.864q7dx2,

i.v. 12108Toxic3.0

±0.4*1.3ToxicToxic2.9

±0.600000000057.5

(40)"7.5

(40)0(10)50.0

(26)43.5(44)17.8(45)

route. Representative data are presented in Fig. 3B. The BR64-DOX conjugate (mole ratio = 3.89) administered i.p. at a doseof 15 mg/kg equivalent DOX on a q2dx5 schedule demonstrated antitumor activity which was equivalent to 11.5TVDDs, and this treatment produced 38% partial tumor regressions. The SN7-DOX conjugate (mole ratio = 5.88), administered at a matching dose and schedule, was not activeagainst L2987 xenografts.

The antitumor activity of BR64-DOX conjugates was compared to that of unconjugated MAb BR64 and mixtures consisting of MAb BR64 and DOX. In these experiments the mixtures were administered either i.p. or i.v. and the DOX dose ofthe mixtures was equivalent to the MTD for the particular routeof administration. As shown in Fig. 4, unconjugated MAbBR64 administered i.p. at a dose of 1000 mg/kg or i.v. at a doseof 500 mg/kg on a q4dx3 schedule did not inhibit the growth ofestablished L2987 tumors. Data presented in Table 1 demonstrate that DOX administered at the MTD by the i.p. route wasnot active against established L2987 tumors. Similarly, mixtures of MAb BR64 (1000 mg/kg) and DOX (4 mg/kg) administered i.p. were not active (Fig. 4A). The i.v. administration ofa mixture consisting of 500 mg/kg MAb BR64 and 8 mg/kgDOX, on a q4dx3 schedule, produced antitumor activity whichwas not significantly different from that of the same dose ofDOX administered alone (Fig. 4B). Taken together, these datademonstrate that any enhancement in antitumor activity observed with BR64-DOX conjugates was not due to synergisticactivity of MAb and DOX. As shown above the antitumoractivity of BR64-DOX conjugates was significantly better than

that of optimized DOX, equivalent doses of nonbinding conjugates, and mixtures consisting of MAb BR64 and DOX.

The antitumor activity of BR64-DOX conjugates was alsocompared to that of optimized DOX in the MCF7 breast carcinoma model. Therapy was initiated 12-14 days after tumorimplantation, when the tumors were 75-125 mm3 in size. Mice

were treated i.v. with optimized DOX (6 mg/kg, q4dx3) or i.p.with BR64-DOX immunoconjugates (15 mg/kg equivalentDOX; mole ratio = 3.89). As shown in Fig. 5, the antitumortumor activity of the BR64-DOX conjugate (5.9 TVDDs) administered at a dose of 10 mg/kg, q2dx4, was superior to thatobserved for optimized DOX (2.5 TVDDs).

The antitumor activity of several preparations of BR64-DOXand SN7-DOX conjugates is summarized in Table 3. The an-titumor activity of BR64-DOX conjugates prepared with moleratios of 3.89-7.85 was consistently superior to that of optimized DOX (Table 1) or equivalent doses of nonbinding SN7-DOX conjugates. The MTD for the conjugates was 30-35mg/kg when administered on a q4dx3 schedule and 15-20mg/kg when administered on a q2dx5 schedule. This is equivalent to a cumulative exposure of approximately 100 mg/kgequivalent DOX. The MTD of the conjugates did not vary as afunction of the mole ratio of the conjugate. The antitumoractivity observed at optimal doses was not significantly differentwhether the conjugate was administered on a q2dx5 (7-13TVDDs) or q4dx3 (9-12 TVDDs) schedule. Although the moleratio of the conjugates was correlated with in vitro potency (Fig.2), a similar relationship was not observed in vivo for conjugatesprepared with mole ratios in the range of 3-8. For example, the

5696

on May 26, 2021. © 1992 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTITUMOR ACTIVITY OF BR64-DOX CONJUGATES

800 -,

600 -

400 -

0 20 40 60

Days Post Implant

80 100

800 -1

600 -

g 400in

200 -

20 40 60 80 100

Days Post ImplantFig. 3. Antitumor activity of DOX and DOX conjugates evaluated against

established L2987 human lung tumor xenografts. A, dose-dependent antitumoractivity of BR64-DOX conjugates (mole ratio = 5.37), compared to that of optimized DOX. Results are from control animals (A) or animals treated with DOXat 8 mg/kg i.v. (•)or BR64-DOX conjugates at 35 mg/kg i.p. (•)or 25 mg/kg i.p.(O) 14, 18, and 22 days after tumor implantation. B, antigen-specific antitumoractivity of BR64-DOX conjugates. The antitumor activity of BR64-DOX conjugates was compared to that of a nonbinding SN7-DOX conjugate. Results arefrom control animals (A) or animals treated i.p. with 15 mg/kg BR64-DOX (moleratio = 5.14) (•)or SN7-DOX (mole ratio = 5.88) (D) 22, 24, 26, 28, and 30 daysafter tumor implantation.

antitumor activity of a BR64-DOX conjugate with a mole ratioof 3.89 (administered q2dx5) was similar to that of a BR64-DOX conjugate with a mole ratio of 7.85 administered on thesame schedule.

DISCUSSION

These experiments have described the antigen-specific activity of BR64-DOX conjugates when evaluated in vitro and inhuman tumor xenograft models. MAb BR64 identifies a Ley-related antigen expressed in high density on carcinomas of several histological types. In addition, the MAb has been shown tobe internalized following antigen-specific binding (29). Becauseof this characteristic of the antibody, immunoconjugates were

produced using an acid-labile hydrazone bond (26, 27). Withthis linker chemistry, internalization of the immunoconjugateinto an acidic environment, such as that of lysosomes or endo-

somes, will result in the release of DOX (26, 33).The ability to deliver a lethal concentration of drug to cells is

dependent upon the density of antigen expressed (i.e., the number of accessible sites for immunoconjugate binding) and thequantity of drug delivered per antibody molecule. Therefore,conjugates prepared with a higher drug:MAb mole ratio shouldbe more potent because more drug is delivered per antibodymolecule bound to target cells. An obvious upper limit to increasing potency by increasing the drug:MAb mole ratio is thereduction in immunoreactivity typically observed with highdrug loading (5, 34). In the present study immunoconjugateswere prepared with drug:MAb mole ratios in the range of 1-8,

1000 -i

800 -

600 -

400 -

200 -

10 20 30 40 50 60

Days Post Implant

800-1

600-

400 -

200-

10 20 50 60

Days Post Implant

Fig. 4. Antitumor activity of MAb BR64, DOX, and mixtures of MAb BR64and DOX evaluated against established L2987 human tumor xenografts. Therapywas administered i.v. (A) or i.p. (B) on days 15, 19, and 23 after tumor implantation. A, treatment administered i.v. A, control; •¿�,DOX, 8 mg/kg; •¿�,MAbBR64, 500 mg/kg; D, a mixture consisting of 8 mg/kg DOX and 500 mg/kgBR64. B, treatment administered i.p. A, control; •¿�.DOX, 4 mg/kg; •¿�MAbBR64. 1000 mg/kg; D, a mixture consisting of 4 mg/kg DOX and 1000 mg/kgMAb BR64.

5697

on May 26, 2021. © 1992 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTITUMOR ACTIVITY OF BR64-DOX CONJUGATES

IODOn

800-

400-

200-

—¿�I—40 60 80

Days Post Implant

Fig. 5. Antitumor activity of BR64-DOX conjugate and optimized DOX evaluated against established MCF7 human breast tumor xenografts. A, control animals; •¿�.DOX administered i.v. at a dose of 6 mg/kg on days 12, 16. and 20 aftertumor implantation: •¿�.BR64-DOX conjugate (mole ratio = 3.89) administeredi.p. at a dose of 15 mg/kg on days 12. 14. 16. and 18 after tumor implantation.

and >90% of immunoreactivity was retained following drugconjugation. As shown above, immunoconjugates producedwith higher mole ratios were significantly more potent in vitrothan were those prepared with lower mole ratios. This relationship was most dramatic for conjugates in which mole ratioswere <3; these conjugates were approximately 10-fold less potent than conjugates prepared with mole ratios of >3. It ispossible that the reduced potency of the low mole ratio immunoconjugates may reflect both the lower drug loading and thepresence of unconjugated BR64; the unconjugated BR64 couldhave competed with BR64-DOX for binding sites on the targetcells and thereby reduced the effective potency of the conjugates. Based upon these in vitro data, immunoconjugates withmole ratios of >3 were selected for further studies.

Immunoconjugates have been constructed using chemother-apeutic agents such as methotrexate (5, 34), Vinca derivatives(7, 35-37), chlorambucil (8), and a variety of anthracyclines

including DOX (4, 22, 28, 30), daunomycin (1, 3), and mor-pholinodoxorubicin (9). In general, the immunoconjugateshave been active in vitro and in vivo. However, antitumor activity was typically evaluated against newly implanted tumors andthe activity was compared to that of equivalent rather thanoptimal doses of the parent compound. The antitumor activityof BR64-DOX conjugates described in this report is comparable or superior to that which has been demonstrated for otherimmunoconjugates. In particular, the activity of BR64-DOXconjugates was evaluated under stringent conditions requiringantigen-specific activity against well established, progressivelygrowing, carcinoma xenografts. The antitumor activity ofBR64-DOX was compared to the maximal activity which couldbe achieved with DOX, the clinically relevant parent compound. The antitumor activity of BR64-DOX conjugates wasclearly superior to that achieved with optimized DOX in boththe L2987 human lung and MCF7 human breast carcinomamodels. The superior efficacy of BR64-DOX conjugates wasevident in terms of both a prolonged delay in tumor growth andan increase in tumor regressions and cures. Complete tumorregressions and cures were observed following treatment withBR64-DOX but were not observed following therapy with optimized DOX. The antitumor activity of the BR64-DOX conjugates was shown to be dose dependent and may be furtherimproved by evaluating the optimal route and schedule for conjugate administration. Although the conjugate mole ratio wascorrelated with in vitro potency, a similar relationship was notobserved in vivo for conjugates prepared with mole ratios in therange of 3-8.

Experiments were performed to ensure that the superior activity of the BR64-DOX conjugate was not due to a carriereffect in which the conjugate served as a depot for the release ofDOX. Nonbinding immunoconjugates were prepared and administered using the same dose, route, and schedule as those forthe binding BR64-DOX conjugates. The BR64-DOX conjugates consistently demonstrated excellent antitumor activity,whereas the nonbinding conjugates were not active against established L2987 tumor xenografts. These data indicate that thesuperior activity of the BR64-DOX conjugates is a result ofantigen-specific binding and is not the result of a pharmacoki-netic change due to slow release of free DOX from the antibody.

Table 3 Antitumor activity of BR64-DOX and nonbinding SN7-DOX conjugates against established L2987 human lung carcinoma xenografts

Optimaldose(mg/kgConjugate,

MR"BR64-DOX.

7.85BR64-DOX,5.37BR64-DOX,5.14BR64-DOX,5.1BR64-DOX,3.89SN7-DOX.

6.35SN7-DOX.5.88SN7-DOX.4.82SN7-DOX.2.57BR64

DOX,7.85BR64-DOX.7.5BR64-DOX.5.37BR64-DOX.4.88SN7-DOX,

4.82SN7-DOX.3.89SN7-DOX.

3.78DOX20151520202015152030203530252030MAb71878776811051448887648877164510773518361732146214492237Schedule*q2dx5q2dx5q2dx5q2dx5q2dx5q2dx5q2dx5q2dx5q2dx5q4dx3q4dx3q4dx3q4dx3q4dx3q4dx3q4dx3TVDD11.66.911.58.612.91.72.31.73.18.910.311.99.23.32.31.0Tumor

regressions(%)Partial600381060000038136350000Complete0000100000130013000Cure2000100000000130000

" MR, mole ratio.AAll therapy was administered by the i.p. route.

5698

on May 26, 2021. © 1992 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTITUMOR ACTIVITY OF BR64-DOX CONJUGATES

Because unconjugated MAbs have been shown to be activeagainst human tumor xenografts, it was possible that immuno- 6

conjugate activity may have been mediated, at least in part, byeffector activities of the M Ab (22, 31, 38, 39). Also, in severalstudies unconjugated mixtures of antibody and drug have beenshown to have greater antitumor activity than does a matcheddose of drug administered alone (22). In the present study, themurine IgGl MAb BR64 was not active against established ghuman tumor xenografts. To address the potential synergisticactivity of DOX and antibody, mixtures of BR64 antibody andDOX were administered by the i.p. or i.v. route at the drugMTD. The activity of mixtures of BR64 antibody and DOX wasnot different from that observed for DOX alone. Therefore, the I0

superior activity of BR64-DOX conjugates reflects the efficacy 11.of the conjugate itself and is not a function of synergistic activity of MAb BR64 and DOX.

To achieve antitumor activity superior to that of DOX it was 12.necessary to administer a cumulative dose of approximately60-90 mg/kg equivalent DOX. It is not clear at this timewhether the high doses of BR64-DOX conjugate required reflect low potency of the conjugate or whether a threshold dose '3-

of MAb must be administered to achieve sufficient accumulation of the BR64-DOX conjugate in the solid tumor xenografts.The conjugates used in these studies were prepared using the 14coupling agent w-succinimidyl 3-(pyridyldithiol)proprionate,which generates a disulfide linkage. Previously, it has beenshown that disulfide-linked ricin A chain conjugates are unstable in vivo and that the use of other linkages, such as a hindereddisulfide, increases the stability of the conjugates (40, 41). Thisincrease in stability likely reflects the reduced susceptibility ofthe hindered disulfide to reduction mechanisms, such as thatinvolving glutathione present in liver and plasma. The low po- 17tency of the BR64-DOX conjugates may result from the instability of the disulfide linker, and the use of more stable linkersmay significantly improve both the potency and the antitumoractivity of BR64-DOX conjugates.

In summary, the studies reported here demonstrate that 19BR64-DOX conjugates mediate antigen-specific cytotoxicity in 2ovitro and are highly effective in inhibiting the growth of established antigen-expressing human tumor xenografts. The anti- 21tumor activity of BR64-DOX conjugates is clearly superior tothat of matching doses of nonbinding conjugate and to optimal 22doses of the relevant clinical compound, DOX.

ACKNOWLEDGMENTS

The authors wish to thank I. Hellstrom and K. E. Hellstrom forgenerously providing the BR64 antibody, A. M. Casazza for helpfulreview of the manuscript, D. S. Field for technical assistance, and K. F.Kadow for determining the binding activity of immunoconjugates.

REFERENCES

IS.

16.

18.

23.

24

25

1. Arnon, R.. and Sela, M. In vitro and in vivo efficacy of conjugates of dauno- 26mycin with anti-tumor antibodies. Immunol. Rev., 62: 5-27, 1982.

2. Hellstrom, K. E., Hellstrom. I., and Goodman, G. E. Antibodies for drugdelivery- In: J. R. Robinson and V. H. L. Lee (eds.). Controlled Drug Delivery, Fundamentals and Applications. Ed. 2, pp. 623-653. New York: Marcel 27Decker, 1987.

3. Dillman. R. O., Johnson. D. E., Shawler, D. L., and Koziol, J. A. Superiorityof an acid-labile daunorubicin-monoclonal antibody immunoconjugate compared to free drug. Cancer Res., 48: 6097-6102. 1988. 28

4. Yang, H. M., and Reisfeld, R. A. Doxorubicin conjugated with a monoclonalantibody directed to a human melanoma-associated proteoglycan suppressesthe growth of established tumor xenografts in nude mice. Proc. Nati. Acad. 29Sci. USA, 85: 1189-1193, 1988.

5. Kulkarni, P. N.. Blair. A. H.. and Chose, T. I. Covalent binding of metho-

5699

trexate to ¡mmunoglobulinsand the effect of antibody-linked drug on tumorgrowth in vivo. Cancer Res., 41: 2700-2706, 1981.Shih. L. B., Sharkey, R. M., Primus, F. J., and Goldenberg, D. M. Site-specific linkage of methotrexate to monoclonal antibodies using an intermediate carrier. Int. J. Cancer, 41: 832-839, 1988.Starling. J. J., Maciak. R. S.. Law, K. L., Hinson, N. A.. Briggs, S. L.,Laguzza, B. C., and Johnson. D. A. In rivo activity of a monoclonal antibody-Vinca alkaloid immunoconjugate directed against a solid tumor membraneantigen characterized by heterogeneous expression and noninternalization ofantibody-antigen complexes. Cancer Res., 51: 2965-2972, 1991.Smyth. M. J., Pietersz, G. A.. Classon, B. J., and McKenzie, I. F. C. Specifictargeting of chlorambucil to tumors with the use of monoclonal antibodies. J.Nati. Cancer Inst., 76: 503-510, 1986.Mueller, B. M., Wrasidlo. W. A., and Reisfeld, R. A. Antibody conjugateswith morpholinodoxorubicin and acid-cleavable linkers. BioconjugatcChem., 7:325-330, 1990.Sands, H. Experimental studies of radioimmunodetection of cancer: an overview. Cancer Res., 50: 809s-813s, 1990.Bernstein, I. D., Eary, J. F., Badger. C. C.. Press, O. W., Appelbaum, F. R.,Martin. P. J., Krohn. K. A., Nelp, W. B., Porter, B., Fisher, D.. Miller, R.,Brown, S., Levy, R., and Thomas, E. D. High dose radiolabelled antibodytherapy of lymphoma. Cancer Res., 50: I017s-I02ls, 1990.Sharkey. R. M., Gold. D. V.. Aninipot, R. V.. Ballance, C.. Newman, E. S.,Ostella, F., Hansen. H. J., and Goldenberg, D. M. Comparison of tumortargeting in nude mice by murine monoclonal antibodies directed againstdifferent human colorectal cancer antigens. Cancer Res., 50: 828s-834s,1990.Thorpe, P. E., Brown, A. N. F., Bremner, J. A. G., Foxwell, B. M. J., andStirpe, F. An immunotoxin composed of monoclonal anti-Thy 1.1 antibodyand a ribosome-inactivating protein from Saponaria offìcinalis:potent anti-tumor effects in vitro and in vivo. J. Nati. Cancer Inst., 75: 151-159, 1985.Vitetta, E. S., Stone. M., Amlot. P., Fay, J., May, R., Till, M.. Newman, J.,Clark, P., Collins, R., Cunningham. D., Ghetie, V., Uhr, J. W., and Thorpe.P. E. Phase I immunotoxin trial in patients with B-cell lymphoma. CancerRes., 51: 4052-4058. 1991.Cazzola, M., Bergamaschi. G., Dezza, L., D'Uva, R., Ponchio. L., Rosti. V.,and Ascari, E. Cytotoxic activity of an anti-transferrin receptor immunotoxinon normal and leukemic human hematopoietic progenitors. Cancer Res., 51:536-541, 1991.FitzGerald, D., Idziorek. T.. Batra, J. K., Willingham, M., and Pastan, 1.Antitumor activity of a thioether-linked immunotoxin: OVB3-PE. Biocon-jugate Chem., /: 264-268, 1990.FitzGerald, D. J., Willingham, M. C., and Pastan, I. Antitumor effects of animmunotoxin made with Pseudomonas exotoxin in a nude mouse model ofhuman ovarian carcinoma. Proc. Nati. Acad. Sci. USA, 83: 6627-6630,1986.Roffler. S. R.. Yu, M. H.. Chen, B. M., Tung. E., and Yeh, M. Y. Therapy ofhuman cervical carcinoma with monoclonal antiboá)-Pseudomonas exotoxinconjugates. Cancer Res., 5/: 4001-4007, 1991.Carter, S. K. Adriamycin: a review. J. Nati. Cancer Inst., 55: 1265-1274,1975.Young. R. C., Ozols, R. F., and Myers, C. E. The anthracycline antineoplas-tic drugs. N. Engl. J. Med., 305: 139-153, 1981.Gallego, J., Price, M. R., and Baldwin, R. W. Preparation of four daunomy-cin-monoclonal antibody 79IT/36 conjugates with anti-tumor activity. Int. J.Cancer. 33: 737-744, 1984.Dillman, R. O., Shawler. D. L., Johnson, D. E., Meyer, D. L., Koziol, J. A.,and Frincke. J. M. Preclinical trials with combinations and conjugates ofT101 monoclonal antibody and doxorubicin. Cancer Res., 46: 4886-4891,1986.Aboud-Pirak. E.. Hurwitz, E., Bellot, F., Schlessinger, J., and Sela, M. Inhibition of human tumor growth in nude mice by a conjugate of doxorubicinwith monoclonal antibodies to epidermal growth factor receptor. Proc. Nati.Acad. Sci. USA, 86: 3778-3781. 1989.Aboud-Pirak, E., Lesur. B., Rao, K. S. P. B., Baurain, R.. Trouet, A., andSchneider, Y. J. Cytotoxic activity of daunorubicin or vindesin conjugated toa monoclonal antibody on cultured MCF-7 breast carcinoma cells. Biochem.Pharmacol., 38: 641-648. 1989.Page, M., Thibeault. D., Noel. C., and Dumas, L. Coupling a preactivateddaunorubicin derivative to antibody. A new approach. Anticancer Res., 10:353-358. 1990.Greenfield, R.. Kaneko, T., Daues, A., Edson, M. A., Fitzgerald, K. A.,Olech, L. J., Grattan, J. A., Spitalny, G. L., and Braslawsky, G. R. Evaluationin vitro of Adriamycin immunoconjugates synthesized using an acid-sensitivehydrazone linker. Cancer Res.. 50: 6600-6607, 1990.Kaneko, T.. Willner, D., Monkovic, I., Knipe, J. O., Braslawsky, G. R.,Greenfield. R. S., and Vyas, D. M. New hydrazone derivatives of Adriamycinand their immunoconjugates: a correlation between acid stability and cytotoxicity. Bioconjugate Chem.. 2: 133-141, 1991.Shih. L. B., Goldenberg, D. M.. Xuan, H., Lu. H., Sharkey. R. M.. and Hall,T. C. Anthracycline immunoconjugates prepared by a site-specific linkage viaan amino-dcxtran intermediate carrier. Cancer Res., 51: 4192-4198, 1991.Hellstrom, I., Garrigues, H. J., Garrigues. U., and Hellstrom, K. E. Highlytumor-reactive, internalizing, mouse monoclonal antibodies to Le'-relatedcell surface antigens. Cancer Res., 50: 2183-2190, 1990.

on May 26, 2021. © 1992 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

ANTITUMOR ACTIVITY OF BR64-DOX CONJUGATES

30. Braslawsky, G. R., Edson, M. A., Pearee, W., Kaneko, T., and Greenfield, R.S. Antitumor activity of Adriamycin (hydrazone-linked) immunoconjugatescompared with free Adriamycin and specificity of tumor cell killing. CancerRes., 50:6608-6614, 1990.

31. Trail, P. A., and Rose, W. C. Development of a human xenograft model forthe evaluation of monoclonal antibody L6-mitomycin immunoconjugates. InVivo. 3: 319-324. 1989.

32. Hart, J. E. Endocrine pathology of estrogens: species differences. Pharmacol.Ther., ¥7:203-218. 1990.

33. Braslawsky, G. R., Kadow, K.. Knipe. J., McGoff. K., Edson, M.. Kaneko, T.,and Greenfield, R. Adriamycin(hydrazone)-antibody conjugates require internalization and intracellular hydrolysis for antitumor activity. Cancer Im-munol. Immunother., 33: 367-374. 1991.

34. Kanellos, J., Pietersz, G. A., and McKenzie, 1. F. C. Studies of methotrexate-monoclonal antibody conjugates for immunotherapy. J. Nail. Cancer Inst.,75:319-329, 1985.

35. Rowland, G. F., Axton, C. A., Baldwin, R. W., Brown, J. P., Corvalan, J. R.F., Embleton. M. J., Gore, V. A., Hellstrom, !.. Hellstrom, K. E., Jacobs, E.,Marsden, C. H., Pimm, M. V., Simmonds, R. G., and Smith, W. Antitumorproperties of vindesine-monoclonal antibody conjugates. Cancer Immunol.Immunother.. 19: 1-7, 1985.

36. Starling, J. J., Maciak, R., Hinson, N. A.. Nichols, C., Briggs. S. L., and

Laguzza, B. C. In vivo efficacy of monoclonal antibody-drug conjugates ofthree different subisotypes which bind the human tumor-associated antigendefined by the KSI/4 monoclonal antibody. Cancer Immunol. Immunother.,28: 171-178, 1989.

37. Johnson, D. A., Baker, A. L., Laguzza, B. C., Fix, D. V., and Gutowski, M.C. Antitumor activity of L/lC2-4-desacetylvinblastine-3-carboxyhydrazideimmunoconjugate in xenografts. Cancer Res., 50: 1790-1794, 1990.

38. Hellstrom, I., Beaumier, P. L., and Hellstrom, K. E. Antitumor effects of L6,an IgG2a antibody reacting with most human carcinomas. Proc. Nati. Acad.Sci. USA, 83: 7059-7063, 1986.

39. Gutowski, M. G., Briggs, S. L., and Johnson, D. A. Epidermal growth factorreceptor-reactive monoclonal antibodies: xenograft antitumor activity aloneor as drug immunoconjugates. Cancer Res., 51: 5471-5475, 1991.

40. Thorpe, P. E., Wallace, P. M., Knowles, P. P., Reif, M. G., Brown, A. N. F.,Watson, G. J., Knyba, R. E., Wawrzynczak, E. J., and Blakey, D. C. Newcoupling agents for the synthesis of immunotoxins containing a hindereddisilludi' bond with improved stability in vivo. Cancer Res., 47: 5924-5931,1987.

41. Worrell, N. R., Cumber, A. J., Parnell, G. D., Mirza, A., Forrester, J. A., andRoss, W. C. J. Effect of linkage variation on the pharmacokinetics of ricin Achain-antibody conjugates in normal rats. Anti-Cancer Drug Design, /: 179-188. 1986.

5700

on May 26, 2021. © 1992 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1992;52:5693-5700. Cancer Res   Pamela A. Trail, David Willner, Shirley J. Lasch, et al.   Tumor Xenograft Models

and in Humanin VitroBR64-Doxorubicin Conjugates Evaluated Antigen-specific Activity of Carcinoma-reactive

  Updated version

  http://cancerres.aacrjournals.org/content/52/20/5693

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/52/20/5693To request permission to re-use all or part of this article, use this link

on May 26, 2021. © 1992 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from