In Vitro and in Vivo Cytotoxicity of an Anti-Osteosarcoma ... · osteosarcoma and some soft tissue sarcomas. INTRODUCTION Adjuvant chemotherapy is a standard treatment approach for

[CANCER RESEARCH 55, 1321-1327, March 15. 1995]

In Vitro and in Vivo Cytotoxicity of an Anti-Osteosarcoma Immunotoxin ContainingPokeweed Antiviral Protein1

Peter M. Anderson,2 Dorothea E. Meyers, Diane E. Hasz, Kristin Covalcuic, Daniel Saltzman, Chand Khanna,

and Fatih M. Uckun

Departments of Pediatrics ¡P.M. A., D. E. H., F. M. U.¡,Therapeutic Radiology ¡D.E. M.. K. C.. F. M. U.I. Surgery ¡D.S.¡,and Small Animal Clinical Sciences ¡C.K.¡.University of Minnesota Biotherapy Program, Minneapolis Minnesota 55455

ABSTRACT

Successful treatment of many patients with osteosarcoma requiresmore effective chemotherapy. Since new agents are needed, we havedeveloped an immunotoxin using TP-3, an IgG2b mAh which recognizes

human and canine osteosarcomas and budding capillaries of tumors. Theplant hemitoxin, pokeweed antiviral protein (PAP), was conjugated toTP-3 to produce an immunotoxin highly active against osteosarcoma.

After 48 h no viable human OHS osteosarcoma cells were present incultures containing TP-3-PAP as demonstrated by the absence of [3H]thy-

midine uptake into DNA. Furthermore, clonogenic assays indicated >3.9log kill of OHS at 18 h. The IC50 of TP-3-PAP against OHS was 3.5 ±1.0(SD) x 10~12 M. TP-3 mAb without PAP had no effect on OHS prolifer

ation; PAP alone had no effect on OHS growth unless concentrations>1000 pM were used. When TP-3-PAP (1.25 ug-10.0 ug) was given i.p. q.d.on days 3-5 after tumor inoculation, a dose-dependent reduction of thenumber of lung métastaseswas observed (/' < 0.001). These results indi

cate that the TP-3-PAP immunotoxin may be useful in the treatment of

osteosarcoma and some soft tissue sarcomas.

INTRODUCTION

Adjuvant chemotherapy is a standard treatment approach for manysarcomas including Ewing's sarcoma, rhabdomyosarcoma, and osteo

sarcoma (1-5). It has been more difficult, however, to demonstrate a

significant benefit of chemotherapy in adult soft tissue sarcomas (6)and osteosarcoma with metastatic disease at diagnosis or after thedevelopment of lung métastases(7). Since the degree of necrosis ofosteosarcoma in the neoadjuvant setting has been shown to be a highlysignificant predictor of disease-free survival (2-3, 5, 8-9), a need

exists to increase the proportion of osteosarcoma patients with favorable initial responses to neoadjuvant chemotherapy, as well as todevelop novel treatments for patients with poor responses (2) ormetastatic disease (5, 7).

The rarity and heterogeneity of sarcomas (10) has made the development of immunotherapeutic agents against these cancers quitedifficult (11). Immunotherapy alone and in combination with chemotherapy after local control has been investigated as a treatmentmodality for canine osteosarcoma. Recent work by MacEwen et al.with muramyl tripeptide phosphatidylethanolamine in liposomes hasdemonstrated improved survival over surgery alone (12) and synergywhen muramyl tripeptide phosphatidylethanolamine liposomes wereused with cisplatin (13).

One promising set of mAbs, TP-1 and TP-3, has been shown to

react with different epitopes of an Mr 80,000 antigen on human andcanine osteosarcoma (14-21). TP-1 and TP-3 also bind a variety ofother human sarcomas including hemangiopericytoma, chondrosar-

Received 10/31/94; accepted 1/16/95.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 with18 U.S.C. Section 1734 solely lo indicate this fact.

1This work was supported by the Children's Cancer Research Fund. The Hedherg

Foundation, and The University of Minnesota Biotherapy Program.2 To whom requests for reprints should be addressed, at Division of Pediatrie Hema-

tology/Oncology, Box 484, University of Minnesota Hospital and Clinic, D557 Mayo, 420Delaware Street SE, Minneapolis MN 55455.

coma, MFH,3 and synovial cell sarcoma (14, 15). Canine osteosar

coma and lung carcinoma were found to react with TP-3 better thanTP-1 (16). The antigen recognized by TP-1 and TP-3 mAbs appearsto be unique among anti-osteosarcoma antibodies. The distribution ofthe TP-l/TP-3 antigen on normal tissues is very limited. Negative

tissues included fibroblasts, peripheral blood cells, cells in the marrow, fetal skin fibroblasts, fetal lung fibroblasts, amniocytes, fibrousconnective tissue, skeletal muscle, cartilage, synovia, peripheralnerve, tonsil, spleen, liver, colon, and lung. Only newly active bonecallus, placenta! endothelial cells, proximal tubule of kidney (weakbinding), and occasional cells in the adrenal medulla were positive forTP-1 and TP-3 (14, 15).

Interestingly, although normal endothelium did not bind TP-1 orTP-3 mAbs, budding capillaries of nonsarcomatous origin were alsoTP-1ATP-3 positive (15). Radioimmunoscintigraphy using 131I-la-

beled TP-1 or TP-3 mAb preparations demonstrated selective tumor

targeting in nude mice (17), dogs with osteosarcoma lung métastases(18), and humans with osteosarcoma métastases(19, 21).

In the case of osteosarcoma, immunotoxin technology may providea potent means to achieve effective therapeutic use of the TP-3 mAb.

Immunotoxins are bifunctional proteins which have been prepared bycovalently linking a cell type-specific mAb to one of a variety ofcatalytic toxins (22-26). A direct comparison of a number of bacterial

and plant toxin immunoconjugates recently showed that PAP constructs were among the most potent immunotoxins tested (25). Favorable preclinical and clinical results have been obtained by our groupagainst lymphocytes and lymphoblasts with PAP containing immunotoxins (26-31). We have applied this experience toward the development of a TP-3-PAP immunotoxin for therapy of osteosarcoma in

dogs and humans. This report describes high in vitro potency ofTP-3-PAP against human osteosarcoma and significant in vivo

efficacy against murine sarcoma lung métastases.

MATERIALS AND METHODS

TP-3 mAb Production and Purification. The TP-3 mAb hybridoma was

obtained from Dr. R. Zalutsky (Duke University Medical Center) with permission of Dr. 0. S. Bruland (The Radium Institute, Oslo, Norway). TP-3

hybridoma cells were cultured in DMEM (Celox, Hopkins, MN) containing 25mM HEPES, 2 mM L-glutamine, 100 units/ml penicillin, 100 /¿g/mlstrepto

mycin, 10 mM nonessential amino acids, 100 mM sodium pyruvate, and 10%PCS (Sigma Chemical Co., St. Louis, MO). BALB/c mice were primed with0.5 ml pristane (Aldrich Chemical Co., Milwaukee, WI) i.p. 7 days beforeinjection of 2 X 10'' TP-3 hybridoma cells i.p. Ascites containing TP-3 mAb

were collected, centrifuged at 12,000 X g for 20 min, pooled, and filteredthrough a 0.22-fi.m filter. TP-3 mAb was further purified using ammonium

sulfate precipitation and affinity chromatography with protein A agarose(Immunopure Plus immobilized protein A; Pierce, Rockford, IL). Elution fromprotein A was accomplished with Immunopure elution buffer (Pierce). TP-3

was dialyzed against PBS and sterile filtered prior to use.Cell Lines. The OHS line is an adherent human osteosarcoma line with

high constitutive expression of the TP-1ATP-3 antigen. OHS was derived by

' The abbreviations used are: MFH. malignant fibrous hisliocytoma; PAP, pokeweed

antiviral protein.

1321

on May 23, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CYTOTOXICITY OF ANTI-OSTI-.OSAKCOMA [MMUNOTOXIN

Postad et al. (32) from an adolescent with metastatic osteosarcoma whichoccurred 13 years after retinoblastoma. For the present studies OHS wasobtained from Dr. Deborah Haines (Western College of Veterinary Medicine,Saskatoon, Saskatchewan, Canada) and passaged in RPMI 1640 with 2 mML-glutamine, 100 units/ml penicillin, 100 /j,g/ml streptomycin, and 10% FCS.

Dl7 (canine osteosarcoma) was obtained from Dr. Stuart Helfand (Universityof Wisconsin, Madison, WI); D17 is negative for TP-3 antigen. The human

CD19+ ALL cell line, RS4;11, was obtained from Dr. John Kersey (University of Minnesota, Minneapolis, MN) and used as a negative control line forTP-3-PAP studies.

TP-3-PAP Immunotoxin Synthesis. TP-3 antibody was produced and

purified using procedures described previously (26). As shown in Fig. 1,TP-3-PAP immunotoxin began to elute approximately 34 min after injection,followed closely by unreacted TP-3 mAb. Free PAP eluted at 56 min and was

well separated from the immunotoxin. The HPLC semipurified material still

contained some unreacted TP-3 mAb. SDS-PAGE scanning of the dried gelrevealed <5% PAP in the final TP-3-PAP immunotoxin preparation which also

contained 14% mAb (M, 150,000), 34% of the M, 180,000 species consistingof 1 PAP molecule disulfide linked to 1 mAb molecule, 34% of the M, 210,000species consisting of 2 PAP molecules linked to 1 mAb molecule, and 18% ofthe Mr 240,000 species consisting of 3 PAP molecules linked to each mAbmolecule (Fig. 1). The absence of significant free PAP contamination in thepurified TP-3-PAP immunotoxin was confirmed by Western blot analysisusing an anti-PAP antibody, as described previously in detail (26).

TP-3-PAP Immunotoxin Activity against Human OHS Osteosarcoma

Cells. Solutions of mAb, toxins, and immunotoxins were tested for effects onOHS cell growth using a [3H]thymidine proliferation assay. Samples were

diluted to appropriate concentrations (between 1 mg/ml and 100 ng/ml ofprotein) in media and added in triplicate to the first row of 96-well flat-bottomed microtiter plates. Samples were then serially diluted 3-fold with the

AkDa

200-

87-

/ B•«TP-3-RAP

RM»

Fig. 1. TP-3-PAP purification and characterization. HPLC elulion profile: TP-3-PAP élûtesat 38

min; free PAP at 56 min. MAB, monoclonal antibody. A, Comassie blue PAGE of TP-3 and TP-3-PAP. B, Western blol with anti-PAP shows no freePAP contamination of TP-3-PAP. C, densitometerprofile of TP-3-PAP. kDa, molecular weight inthousands.

0.2000 0.1400 0.0800 0.0200

Absorbance (560nm)

-0.040 -0.100

1322

on May 23, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CYTOTOXICITY OF ANTI-OSTEOSARCOMA IMMUNOTOXIN

use of a multichannel pipettor apparatus by adding 50-fil sample from row Ato 100-/xl media in row B, mixing, and repeating the procedure until the entire

plate was serially diluted. OHS cells were grown to confluence and removedby brief treatment with 0.5% trypsin with EDTA (Sigma) for 5 min at 37°C,washed twice in media, and adjusted to a concentration of 8 x IO4 cells/ml

before adding 0.05 ml to each well of the 96-well microtiter plate. Plates wereincubated in a 5% CO2 atmosphere at 37°Cfor 48-96 h. Cell growth was

monitored using an inverted microscope.After incubation of samples and indicator cells for 2-4 days, 25 fj.1(2 fiCi)

of [3H]thymidine (DuPont New England Nuclear, Boston, MA) was added to

each well, and plates were incubated for 6 h prior to harvesting DNA onto filterpaper discs with a PHD cell-harvesting apparatus (Cambridge Technology,

Inc., Watertown, MA). After addition of liquid scintillation fluid (Cytoscint;1CN Biochemicals, Costa Mesa, CA), radioactivity was determined using anLKB 1216 liquid scintillation counter. Data was analyzed using an Excelmacro routine written by Dr. Bob Jarvis (Department of Computer Sciences,University of Minnesota, Minneapolis, MN) to determine the mean and SD ofeach triplicate set of samples. Clonogenic assays were done with OHS with theuse of methods reported previously (33).

In Vivo Use of TP-3-PAP. Mice were fed and housed by University of

Minnesota Research Animal Resources in accordance with NIH guidelines.Procedures and protocols involving live animals were approved by the University of Minnesota Animal Care Committee. The MCA106 sarcoma (34) wasobtained from Dr. Jim Mule' (National Cancer Institute, Bethesda, MD) and

serially passaged in female C57BL/6 mice. Tumors were harvested, minced,and digested by stirring on a magnetic stir plate for 4 h using 0.4 mg/mlhyaluronidase, 0.05 mg/ml DNase, and 4.0 mg/ml collagenase (Sigma) inRPMI 1640 with 100 units/ml penicillin, 100 jacg/ml streptomycin, and 2 mML-glutamine. Cells were filtered with Cell Strainers (Falcon; Becton DickinsonLabware, Franklin Lakes, NJ), washed 3 times in HBSS without Ca2+ orMg2+, and concentration adjusted to 1 X IO5 cells/ml. Pulmonary métastases

20000l

fO

So.o

og

>.

50000-

40000-

30000-

20000-

10000

10'1 10° 101 „10* 103 10

HSA (control)TP-3 MABAntl-CD19 PAP

TP-3-PAP

5

pM10

50000

ntl-CD19-PAP

PAP (no MAB)

10'210"110° 101 10

pM10 10 10

Fig. 2. Proliferation of human OHS osteosarcoma in relation to TP-3 mAb, PAP, andTP-3-PAP immunotoxin. Selective and significant inhibition of [3H]thymidine uptake into

DNA was observed when the TP-3+ human osteosarcoma cell line was cultured for48-96 h in the presence of TP-3-PAP. A, TP-3-PAP, but not TP-3 mAb alone, eliminatedgrowth of OHS. HSA, human serum albumin. B, The TP-3-binding moiety was necessaryfor the effect of PAP toxin on OHS at low concentrations. B43-PAP, an anti-CD19-PAPimmunotoxin which does not bind CD19-negative OHS cells, had similar effects as freePAP at very high concentrations (>1000 pM).

D17 osteosarcoma (TP-3 negative)

HSAAntl-CDI »-PAPTP-3 MABTP-3-PAP

18000-

si¡O 14000-

lii o•

10000-

6000

RS4;11 CD19TP3- cell line

HSATP-3 MAB

TP-3-PAP

PAP

Anti-CD! 9-PAP

10° 101 10'

PM

10' 10'

Fig. 3. Effect of TP-3-PAP on TP-3-negative cell lines. A, effect of culture on canineD17 osteosarcoma in the presence of human serum albumin (HSA), TP-3 mAb, TP-3-PAPimmunotoxin or B43-PAP (Anti-CDl9-PAP) immunotoxin. B, effect of human serumalbumin, TP-3 mAb, TP-3-PAP, PAP, and anti-CD 19-PAP (B43-PAP) on the CD19+RS4;11 human ALL cell line. Inhibition at low concentrations requires specific binding ofthe mAb moiety.

were established by i.v. injection of MCA106 sarcoma cells (0.4 ml containing40,000 cells/mouse) into the tail vein of 6-8-week-old female C57BL/6 mice.

Groups of 10 mice with pulmonary métastaseswere treated with antibodyalone or immunotoxin preparations i.p. The number of métastaseswas evaluated by direct counting 14 or 17 days after establishment of métastases.Afterasphyxiation with CO,, India ink (5% with 3 gtt NH4OH/1()0 ml) was injectedinto the trachea. Lungs and heart were removed en bloc and placed intoFekete's solution (300 ml 70% ethanol-30 ml 10% formalin-15 ml glacial

acetic acid). Lungs were coded and counted by at least two blinded observers.Differences in the number of métastasesbetween treatment groups wereevaluated using Student's unpaired i test; differences in weight of individualmice on days 0 and 7 after tumor inoculation were compared using Student's

paired t test (InStat; GraphPad Software, San Diego CA).

RESULTS

TP-3-PAP Immunotoxin Efficiently Kills TP-3+ OHS SarcomaCells. Fig. 2 shows the effect of TP-3 mAb, TP-3-PAP, PAP alone,

and an irrelevant immunotoxin construct which binds CD19 on Bcells, B43-PAP, on proliferation of human OHS osteosarcoma cells.The TP-3 mAb alone (i.e., without PAP toxin) had no effect onproliferation; cells incorporated [3H]thymidine into DNA in a manner

identical to media with human serum albumin (Fig. 2A). TP-3-PAP,however, completely eliminated uptake of [3H]thymidine in the first 4

wells which had OHS cells; OHS did not survive immunotoxintreatment until TP-3-PAP was diluted to 20 pM or less.

Different lots of TP-3-PAP yielded reproducible and highly efficient killing of OHS. [3H]Thymidine proliferation assays using OHS

in 5 separate experiments with lot 1 of TP-3-PAP yielded a mean 1C5(I

value of 3.1 ±1.0 pM. Three different experiments using a second lotof TP-3-PAP yielded a mean IC50 of 4.1 ±0.3 pM. The overall mean

IC50 was 3.5 ±1.0 pM.Killing of Cells by TP-3-PAP Is Highly Specific for Cells Ex

pressing TP-3 Antigen. PAP alone or B43-PAP, an anti-CD19 im

munotoxin, had no effect on OHS proliferation until concentrationswere 10,000 pM or more (Fig. 2B). This represents a >3000-foldincrease in cytotoxicity if the TP-3 mAb was conjugated to the PAPmoiety. If tumors did not express the TP-3 antigen, killing by TP-3-

1323

on May 23, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CYTOTOXICITY OF ANTI-OSTEOSARCOMAIMMUNOTOXIN

Table I Antitumor activity of TP-3-PAP against clonogenic osteosarcoma cells (OHS human osteosarcoma clonogenic assay)

Incubation TP-3-PAP concentrationlime (h)(ng/ml)4

041(143041004300410004

300018018118101830181001830018100018

3000Colony

units(mean ±SEM)4588

±15564588±15562683

±9102683±910313

±1354±14±14588

±15562052±550917+311313

±321120±13562

+274±14+

1Log

kill(Mean ±SEM)0.00

±0.000.00±0.200.23+0.200.23+0.201.16

±0.233.91±0.233.91+0.230.00±0.000.34±0.220.69±0.200.58+0.181.16

±0.231.86+0.233.91+0.233.91

+ 0.23

PAP did not occur at concentrations <104 pM (Fig. 3, A and 0).

B43-PAP, however, was active against the CD19+ cell line RS4;11

(Fig. 3B). Thus, killing by the PAP immunoconjugates was conferredby specific mAb binding.

We have used a highly sensitive in vitro serial dilution clonogenicassay system to determine the log kill efficacy of TP-3-PAP immu-

notoxin against clonogenic OHS human osteosarcoma cells. As shownin Table 1, a 4-h treatment with 100-3000 ng/ml TP-3-PAP at37°C/5%CO2 killed clonogenic OHS cells in a dose-dependent fash

ion with a maximum of 3.9 ±0.2 logs at 1000 ng/ml (5.6 nM).Notably, this 4-h treatment protocol with TP-3-PAP concentrations

£100 ng/ml did not significantly inhibit the clonogenic growth ofOHS cells (log kill ^ 0.2 log). By comparison, an 18-h exposure to1-3000 ng/ml TP-3-PAP killed clonogenic OHS cells in a dose-

dependent fashion with 1.2 log kill at 100 ng/ml and >3.9 logs kill atconcentrations >300 ng/ml (Table 1).

TP-3-PAP Is Active in Vivo against Lung Métastases.Table 2and Fig. 4 summarize results using TP-3-PAP in mice bearingMCA 106 pulmonary métastases.Neither TP-3 mAb alone nor irrelevant B43-PAP immunotoxin had any effect on the number of lungmétastases(Table 2). Three consecutive days of i.p. TP-3-PAP immunotoxin treatment, however, was able to significantly red'ice numbers of pulmonary métastases(P < 0.04). A dose-response relation

ship was examined in a second experiment (Fig. 4). Reduction ofpulmonary métastasesby TP-3-PAP was dose related and highly

significant at doses tested (Table 2). Interestingly, not only weresignificantly fewer métastasesseen in TP-3-PAP treated mice, but thesize of lung métastasesin TP-3-PAP-treated mice was much smaller

than métastasesin control mice (data not shown). Cumulative doses ofTP-3-PAP required for significant antitumor effects were between3.75 and 30 jag/mouse (0.2-1.5 mg/kg).

Toxicity of TP-3-PAP in Mice. Groups of mice receiving 10.0 ^gTP-3-PAP immunotoxin 3 times every day had the greatest observable

effects on activity. These mice also failed to gain weight during thedose-response experiment (Fig. 5/4). Mice treated with 2.5 or 1.25 fig

three times every day appeared healthy and had normal weight gain.Effects of the less dose-intense but more prolonged schedule of 5.0 /¿gTP-3-PAP i.p. five times every day were also significant (Fig. 5ß).

Although the activity of this group was nearly normal, the effect of 5days of TP-3-PAP therapy on weight was highly significant (r test,P = 0.0001).

DISCUSSION

PAP, which belongs to the class of plant hemitoxins includinggelonin, saporin, and momordica charanthia inhibitor, is one of themost active ribosomal inactivating proteins. In a comparison of

cytotoxicity of anti-mouse IgG immunotoxins gelonin, ricin A chain,

momordin, dianthin 32, saporin, and PAP had ICM>estimates of 1000,500, 20.0,10.7,5.5, and 2.6 pM,respectively (25). TP-3-PAP had an IC50of 3.5 pMwhich is similar to the in vitro potency of the PAP anti-mouse

IgG immunotoxin tested by Bolognesi et al. (25). Immunotoxins withIC50values of less than 100 pMare considered excellent. Thus, TP-3-PAPhas very high in vitro cytotoxic potency. Since we have seen steady-stateconcentrations of 500-1000 ng/ml achieved safely in clinical trials ofanti-CD19-PAP,4 the ability of TP-3-PAP to kill 3.9 logs of clonogenic

OHS at 300 ng/ml strongly supports a hypothesis that therapeutic exposure levels to TP-3-PAP can be achieved /'/;vivo in dogs or humans with

osteosarcoma without excessive toxicity due to the PAP moiety.The activity of an immunotoxin depends not only on the toxin

utilized but also on efficient binding of antibody to antigen, endocy-tosis, and intracellular release of functional ribosome-inactivating

proteins. Since the potency of PAP is such that a few molecules in thecytoplasm are probably sufficient to kill a cell, TP-3 antigen density

may be less important than specificity of binding in determining theultimate usefulness and therapeutic index of this particular immunotoxin. Bruland and Phil (19) recently summarized the current state ofknowledge of distribution of the TP-1/TP-3 antigen on normal tissues

and mesenchymal tumors. Osteosarcomas stain intensely positive atthe surface and TP-l/TP-3 staining of osteosarcomas is homogenous

in all regions of the tumors. Significant heterogeneity, however, wasseen in the TP-l/TP-3 antigen expression of soft tissue sarcomas. For

example, 4 of 11 MFH specimens were strongly positive, 4 of 11MFH specimens were weakly positive, and 3 of 11 MFH werenegative for TP-l/TP-3 binding.

The very limited tissue distribution of the TP-3 antigen makes it anattractive choice for future in vivo immunotoxin studies. The TP-3

antibody recognizes an epitope present on selected dog cancers including osteosarcoma and lung carcinoma. Thus studies of TP-3-PAP

in dogs with spontaneous occurring osteosarcomas may be useful indetermining in vivo antitumor efficacy in a relevant tumor modeland whether unusual toxicities related to the binding of endothelialcells of neovasculature may be a problem. In humans it appearsthat only placenta! endothelium and the budding capillaries oftumors have TP-l/TP-3 antigen; resting endothelial cells do not

stain positively (15).Radioimmunoscintography in dogs and man with I3'l-labeled TP-1

F(ab')2 showed accumulation at clinical and occult osteosarcoma

métastases(18-21). Studies of radiolabeled TP-3 in osteosarcoma

nude mice with human tumors showed 16% of the dose/g tissue wasfound in the tumor after 24 h (17). Tumor:blood ratios of 3, 4, and 6

4 F. M. Uckun, unpublished observations.

1324

on May 23, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CYTOTOXICITY OF ANTI-OSTEOSARCOMA [MMUNOTOXIN

were seen 24, 48, and 72 h after injection of TP-3-IgG, respectively.Since tumonblood ratios after TP-3 administration remained morethan or equal to 1.0, repeated doses of the TP-3 mAb could possibly

accumulate in osteosarcomas. However, since osteosarcoma is a relatively radioresistant cancer, curative therapy using radioconjugatesof TP-3 may be difficult to achieve.

Repeated doses or therapeutic courses of immunotoxins could beproblematic because of immunogenicity of the toxin moiety. Recentexperience with B43-PAP immunotoxin in patients with ALL indi

cates that this preparation is one of the least immunogenic preparations with a small minority of patients having either serious humananti-mouse antibodies or human anti-PAP antibodies. Whether this is

due to the underlying disease (acute lymphoblastic leukemia), theanti-B-cell effect of the B43-PAP immunotoxin, use of cyclophosph-

amide with the immunotoxin, or low immunogenicity of the PAPprotein remains to be determined. Future studies with TP-3-PAP inlarger animals such as dogs may determine whether TP-3-PAP also

has low immunogenicity.Good therapeutic results were obtained using TP-3-PAP against a

TP-3+ murine soft tissue sarcoma lung métastasesin vivo. The

efficient decrease of numbers of lung métastasesmay possibly beaccounted for by several different mechanisms: (a) tumor destructionand subsequent induction of an efficient cellular immune response tothe sarcoma cells may eliminate some pulmonary métastases;(b)

Table 2 Reduction of lung métastasesafter TP-3-PAP but no effect of irrelevantiminunoiitxin B43-PAP or TP-3 MAB ahne

No. oflungTreatment*HBSS

(control)TP-3MABTP-3-PAPTP-3-PAPTP-3-PAPB43-PAPDose

ofimmunotoxi^None

0.01.13.310.010.0métastases"Mean10.2016.405.202.200.756.50SEM2.502.801.704.200.752.80Student's

r testPvalue1*NSNS0.0470.039NS

" Métastasescounted on day 14.' MCA 106 sarcoma has no CD19 epitopes recognized by B43-PAP immunotoxin;

however MCA 1(16tumor cross-reacts with TP-3 MAB.' Measured in ng/day given on days 3, 4. and 5.

Compared to control group. NS. not significant.

60 i

Mlcroarams TP-3-PAP/dav

Fig. 4. Dose response of TP-3-PAP immunotoxin against MCA106 sarcoma lungmétastases.Increasing doses of TP-3-PAP on days 2. 3, and 4 resulted ¡nsignificantlyfewer pulmonary métastasesthan in control mice 17 days after establishing lung métastases (/ test, P = 0.0001 for 5 and 10 /Ag groups). Columns, mean; bars, SD.

0 wtdayto•wt. day 12

15010 5 2.5 1.25 0

meg TP-3-PAP (ip qd x 3)

ES

£oí

I

• Control0 TP-3-PAP

0 7Day of Experiment

Fig. 5. Effect of TP-3-PAP therapy on weigh! (HI.) gain of mice. A. total weight ofmice (n = 10/group) on days 10 and 12 after Iherapy with TP-3-PAP on days 3. 4, and5. B. individual weights of mice in an experiment in which mice in the TP-3-PAP groupreceived 5 ^tg immunotoxin i.p. on days 3-7. The control had no change in weight(Student's paired / test, P = 0.85). However, mice receiving TP-3-PAP had a mean lossof 1.7 g (-9% body weight; Student's paired t test, P = O.(KX)l).

induction of a cellular immune response against the TP-3 ligand could

direct the immune response to the tumor; and (c) a direct effect ofimmunotoxin on the neovasculature of the sarcoma métastasesispossible.

It is possible that TP-3-PAP may also act by action on tumor

neovasculature. Folkman and others have elegantly reviewed thecurrent state of knowledge of tumor angiogenesis including dataindicating dependence of tumors >1 mm in diameter for angiogenesis(35-39). Recent studies, using a murine model with immunotoxin

against MHC class II which is on tumor vasculature and anotherimmunotoxin against MHC class I on neuroblastoma tumor cells,demonstrated synergy of the two immunotoxins (40, 41). Thus, tumorvascular targeting by immunotoxins may significantly increase efficacy. Since TP-3 antigen is present on budding capillaries of a widevariety of tumors (15, 16), it is possible that the TP-3-PAP immuno

toxin may inhibit the growth of murine lung métastasesby selectivelydestroying tumor vascular endothelium as well as the minority ofsarcoma cells which bear the TP-3 antigen.

The limitations of mAb therapy and immunotoxin therapy ofcancer are many (42-45). These include low specific uptake bytumor (42), toxicity to "innocent bystander" cells which bind the

mAb or toxin, physiological barriers such as the relatively tightendothelium of the lung compared to liver and spleen (43) orincreased interstitial pressure within tumors (44), and the production of antibodies to mAb and/or toxins. Although PAP immunotoxins may be less immunogenic than ricin immunoconjugates,human anti-mouse antibodies and human anti-PAP antibodies havebeen seen in some patients with ALL.5 Therefore, strategies to

1F. M. Uckun, unpublished data.

1325

on May 23, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CYTOTOXICITY OF ANTI-OSTEOSARCOMA IMMUNOTOXIN

reduce the occurrence of human anti-mouse and anti-PAP antibodies such as "humanization" of the TP-3 antibody, concurrent use of

cyclophosphamide, and/or induction of split tolerance (e.g., usinga combination of cyclophosphamide, antigen, and IgG; Ref. 46)may become important if multiple courses or prolonged administration of TP-3-PAP are needed for durable results.

At least one other immunotoxin which binds osteogenic sarcomahas been reported, 791T/36-RTA (47). Since 791T/36 mAb bindscolorectal, gastric, and ovarian cancer cells, but TP-3 does not, theTP-3-PAP immunotoxin in our report may have different biodis-tribution characteristics or mechanisms of internalization than does791/36-RTA (48). Our preliminary results in the context of previous work on the tissue and tumor distribution of TP-3 antigenindicate that TP-3-PAP has high potential to become a novel andeffective new agent against osteosarcomas and possibly some softtissue sarcomas. Although TP-3-PAP immunotoxin may possiblyfacilitate its own penetration into tumors by binding the buddingcapillaries of these tumors, further studies to investigate toxicityand mechanism(s) of action of TP-3-PAP and efficacy againstsarcomas are warranted.

ACKNOWLEDGMENTS

We gratefuly acknowledge Dr. 0yvind S. Bruland for his provision ofthe TP-3 hybridoma to make these studies possible. P. M. A. acknowledgesthe secretarial assistance of Cyndie Symanetz for the typing of references,Theresa Eisenkraft and Dr. Marcela Zebede for cell culture work, and Drs.William Woods and Jeffry Klausner for their encouragement to undertakethis project.

REFERENCES

1. Balis, F. M., Holccnberg, J. S., and Poplack, D. G. General principals ofchemotherapy. In: P. A. Pizzo and D. G. Poplack (eds.), Principles and Practiceof Pediatrie Oncology, Ed. 2, pp. 197-245. Philadelphia: J. B. LippincottCompany, 1993.

2. Meyers, P. A., Heller, G., Hcaley, J., Huvos, A., Lane, J., Marcove, R., Applewhite,A., Vlamis. V.. and Rosen, G., Chemotherapy for nonmetastatic osteogenic sarcoma:the Memorial Sloan-Keltering experience. J. Clin. Oncol., 10: 5-15, 1992.

3. Hudson, M., Jaffe, M. R., Jaffe, N., Ayala, A., Raymond, A. K., Carrasco, H.,Wallace, S., Murray. J., and Robertson, R. Pediatrie osteosarcoma: therapeutic strategies, results, and prognostic factors derived from a ID-year experience. J. Clin.Oncol., 8: 1988-1997, 1990.

4. Frei, E., Ill, and Goorin, A. Osteogenic sarcoma: the development of curativetreatment. In: J. F. Novak and J. M. McMaster (eds.). Frontiers of OsteosarcomaResearch, pp. 5-13. Bern, Switzerland: Hogrefe & Hubcr Publishers, 1993.

5. Ward, W. G., Mikaelian. K., Dorey, F., Mirra, J. M., Sassoon, A., Holmes, E. C,Eilher, F. R., and Eckhardt, J. J. Pulmonary métastasesof stage IIB osteosarcoma andsubsequent pulmonary métastases.J. Clin. Oncol., 12: 1849-1858, 1994.

6. Elias, A. D., and Amman, K. H. Adjuvant chemotherapy for soft tissue sarcoma:an approach in search of an effective regimen. Semin. Oncol., 16: 305-311,

1989.7. Meyers, P. A., Heller, G., Healey, J., Huvos, A., Applewhite, A., Sun, M., and

LaOuaglia, M. Osteogenic sarcoma with clinically detectable metastasis at initialpresentation. J. Clin. Oncol., //: 449-453, 1993.

8. Raymond, A. K., Chawla, S. P., Carrasco, C. H., Ayala, A. G., Fanning, C. V., Orice,B., Arman, T., Plager, C., Papadopoupolos, N. E. J., Edeiken, J., Wallace, S., Jaffee,N., Murray, J. A., and Benjamin, R. S. Osteosarcoma chemotherapy effect: aprognostic factor. Semin. Diagnostic Pathol., 4: 212-236, 1987.

9. Winkler, K., Beron, G., and Detting, G. Neoadjuvant chemotherapy of osteosarcoma:results of a randomized cooperative trial (COSS-82) with salvage chemotherapybased on histologie tumor response. J. Clin. Oncol., 6: 329-337, 1988.

10. McClay, E. F. Epidemiology of bone and soft tissue sarcomas. Semin. Oncol., 16:264-272, 1989.

11. McClay, E. F., and Slovin, S. F. Immunotherapeutic approaches to the treatment ofbone and soft tissue sarcomas. Semin. Oncol., 16: 328-332, 1989.

12. MacEwen, E. G., Kurzamn, I. D., and Rosenthal, R. C. Therapy for osteosarcoma indogs with intravenous injection of liposome encapsulated muramyl tripeptide. J. Nati.Cancer Inst., SI: 935-938, 1989.

13. MacEwen, E. G., Kurzman, 1. D., Rosenthal, R. C., Fox, L., and Madewell, B. R.Combined liposome-encapsulated muramyl tripeptide and cisplatin in dogs withosteosarcoma. In: J. F. Novak and J. H. McMaster (eds.), Frontiers of OsteosarcomaResearch, pp. 117-119. Bern, Switzerland: Hogrefe & Huber Publishers, 1993.

14. Bruland, 0., Fodstad, 0., Funderud, S., and Pihl, A. New monoclonal antibodiesspecific for human sarcomas. Int. J. Cancer, 38: 27-31, 1986.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

1326

Bruland, 0. S., Fodstad, 0., Stenwig, A. E., and Pihl, A. Expression and characteristics of a novel human osteosarcoma-associated cell surface antigen. Cancer Res.,48: 5302-5309, 1988.

Haines, D. M.. and Bruland, 0. S. Immunohistochemical detection of osteosarcoma-associated antigen in canine osteosarcoma. Anticancer Res., 9: 903-908,

1989.Bruland, 0., Fodstad, 0., Skretting, A., and Pihl, A. Selective localisation of tworadiolabelled anti-sarcoma monoclonal antibodies in human osteosarcoma xenografts.Br. J. Cancer, 56: 21-25, 1987.

Haines, D. M., Bruland, 0. S., Matte, G., Wilkinson, A. A., Meric, S. M., and Fowler,J. D. Immunoscintigraphic detection of primary and metastatic spontaneous canineosteosarcoma with F(ab')2 fragments of osteosarcoma-associated monoclonal anti

body TP-1. Anticancer Res., 12: 2151-2158, 1992.

Bruland, 0. S., and Pihl. A. Immunoscintigraphy and radioimmunotherapy: usefulapproaches in the management of osteogenic sarcoma. In: J. F. Novak and J. H.McMaster (eds.). Frontiers of Osteosarcoma Research, pp. 149-159. Bern, Switzer

land: Hogrefe & Huber Publishers, 1993.Page, R. L., Garg, P. K., Garg, S., Archer, G. E., Bruland, 0. S., and Zalutsky, M. R.PET imaging of osteosarcoma in dogs using fluorine-18 labeled monoclonal antibodyFab fragment. J. Nucí.Med., 35: 1506-1513, 1994.

Bruland, 0. S., Fodstad, 0., Solheim, 0. P., Skretting, A., Winderen, M., Michaelsen,T., and Phil, A. Immunoscintography of bone sarcomas: results in 5 patients. Eur. J.Cancer, 30: 1484-1489, 1994.

Thorpe, P. E., and Ross, W. The preparation and cytotoxic properties of antibodytoxin conjugates. Immunol. Rev., 62: 119-158, 1982.Pastan, I., Willingham, M. C., and Fitzgerald, J. P. Immunotoxins. Cell, 47: 641-648,

1986.Vitella, E. S., Fulton, R. J., May, R. D., Till, M. and Uhr, J. W. Redesigning nature's

poisons to create anti-tumor reagents. Science (Washington DC), 238: 1098-1104,

1988.Bolognesi, A., Tazzari, P. L., Tassi, C., Gromo, G., Gobbi, M., and Stirpe, F. Acomparison of anti-lymphocyte immunotoxins containing different ribosome-inacti-vating proteins and antibodies. Clin. Exp. Immunol., 89: 341-346, 1992.

Myers, D. E., Irvin, J. D., Smith, R. S., Kuebelbeck, V. M., and Uckun, F. M.Production of a pokewced antiviral protein (PAP) containing immunotoxin, B43-PAP, directed against the CD 19 human B lineage lymphoid differentiation antigen inhighly purified form for human clinical trials. J. Immunol. Methods, 136: 221-238,1991.Uckun, F. M., Chclstrom, L. M., Irvin, J. D., Finncgan, D., Günther,R„Young, J.,Kuebelbeck, V., Myers, D. E., and Houston, L. L. In vivo efficacy of B43 (anti-CD 19)-pokeweed antiviral protein immunotoxin against BCL-1 murine B-cellleukemia. Blood, 79: 2649-2661, 1992.Uckun, F. M., Chelstrom, L. M., Finncgan, D., Tucl-Ahlgrcn, L., Manivel, C.,Irvin, J. D., Myers, D. E., and Günther,R. Effective immunochemotherapy ofCALLA + CU+ human pre-B acute lymphohlastic leukemia in mice with severecombined immunodeficiency using B43 (anti-CD19) pokewecd antiviral proteinimmunotoxin plus cyclophosphamide. Blood, 79: 3116-3129, 1992.

Uckun, F. M., Manivel, C., Arthur, D., Chelstrom, L., Finncgan, D., Irvin, J.,Tucl-Ahlgrcn, L, Myers, D. E., and Günther,R. In vivo efficacy of B43 (anti-CD19)-pokcwccd antiviral protein immunotoxin against human prc-B acute lymphoblasticleukemia in mice with severe combined immunodeficiency. Blood, 79: 2201-2214,1992.Günther,R., Chelstrom, L. M., Finnegan, D., Tuel-Ahlgren, L., Irvin, J. D., Myers,D. E., Uckun FM. In vim anti-leukemic efficacy of G3.7 (anti-CD7)-pokeweedantiviral protein immunotoxin against human T-lineage acute lymphoblastic leuke-mia/Iymphoma in mice with severe combined immunodeficiency. Leukemia, 7:298-309, 1993.

Uckun, F. M., and Reaman, G. H. Immunoloxins for treatment of leukemia andlymphoma. Leukemia Lymphoma, in press, 1995.Fodstad, 0., Brogger, A., Bruland, 0., Solheim, O., Nesland, J., and Pihl, A.Characteristics of a cell line established from a patient with mulitple osteosarcoma,appearing 13 years after treatment for bilateral retinoblastoma. Int. J. Cancer, 38:33-40, 1986.Uckun, F. M., Gajl-Peczalska, K. J., Kersey, J. H., Houston, L. L., and Vallera, D. A.

Use of a novel colony assay to evaluate the cytotoxicity of an immunotoxin containing pokeweed antiviral protein against blast progenitor cells freshly obtained frompatients with common B-lineage acute lymphoblastic leukemia. J. Exp. Med., 163:347-368, 1986.Rosenberg, S. A., Shu, S., Schwartz, S. L., and Rosenberg, S. A. Adoptive immu-notherapy of established pulmonary métastaseswith LAK cells and recombinantinterleukin-2. Science (Washington DC), 225: 1487-1489, 1984.

Godal. A.. Kumle, B., Phil, A., Juell, S., and Postad, 0. Immunotoxins directedagainst the high-molecular weight melanoma assosiated antigen. Identification ofpotent antibody-toxin combinations. Int. J. Cancer, 52: 631-635, 1992.

Folkman, M. J. Antiangiogenesis. In: V. T. DeVita, Jr., S. Hellman, and S. A.Rosenberg (eds.), Biologic Therapy of Cancer, pp. 743-753. Philadelphia: J. B.Lippincott Company, 1991.Dvorak, H. F., Nagy, J. A., and Dvorak, A. M. Structure of solid tumors and theirvasculature: implications for therapy with monoclonal antibodies. Cancer Cells, 3:77-85, 1991.

Denekamp. J. Vascular attack as a therapeutic strategy for cancer. Cancer MetastasisRev., 9: 267-282, 1990.Olson, K. A., French, T. C., Vallee, B. L., and Fett, J. W. A monoclonal antibody tohuman angiogenin supresses tumor growth in athymic nude mice. Cancer Res., 54:4576-4579, 1994.

Burrows. F. J., Watanabe, Y., and Thorpe, P. A murine model for antibody directed

on May 23, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CYTOTOXICITY OF ANTI-OSTEOSARCOMA IMMUNOTOXIN

targeting of vascular endothclial cells in solid tumors. Cancer Res., 52: 5954-5962, uptake in solid tumors: role of plasma kinetics, capillary permeability, and binding.1992. Cancer Res., 50: 7382-7392. 199(1.

41. Burrows, F. J., and Thorpe, P. E. Eradication of large solid tumors in mice with an 46. Nilsson, I. M., Berntorp. E., and Zettervall, O. Induction of split tolerance and cinica!immunotoxin directed against tumor vasculature. Proc. Nati. Acad. Sci. USA. 90: cure in high-responding hemophiliacs with factor IX antibodies. Proc. Nati. Acad.8996-9000, 1993. Sci. USA, 83: 9169-9173, 1986.

42. Sands. H. Radioimmunoconjugates: An overview of problems and promises. Anti- 47. Embleton. M. J., Byers, V. S., Lee, H. M., Scannen, P., Blackhall, N. W., andbody Immunoconj. Radiopharmaceut., /: 213-226. 1988. Baldwin. R. W. Sensitivity and selectivity of ricin toxin A chain-monoclonal anilbody

43. Kennel, S. J., Falcioni. R., and Wesley. J. W. Microdistribution of specific rat 791T/36 conjugates against human tumor cell lines. Cancer Res.. 46: 5524-5528,monoclonal antibodies to mouse tissues and human tumor xenografts. Cancer Res., 1986.51: 1529-1536. 1991. 48. Byers, V. S.. Pawluczyk, I. Z. A.. Hooi, D. S. W., Price, M. R., Carroll, S., Embleton.

44. Jain, R. K. Determinants of tumor blood flow: a review. Cancer Res., 48: 2641-2658, M. J., Garnett. M. C, Berry, N., Robins, R. A., and Baldwin. R. W. Endocytosis of1988. immunotoxin 791T/36-RTA by tumor cells in relation to its cytotoxic action. Cancer

45. Sung, C., Youle, R. J., and Dedrick, R. L. Pharmacokinetic analysis of immunotoxin Res., 51: 1990-1995, 1991.

1327

on May 23, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1995;55:1321-1327. Cancer Res   Peter M. Anderson, Dorothea E. Meyers, Diane E. Hasz, et al.   Immunotoxin Containing Pokeweed Antiviral Protein

Cytotoxicity of an Anti-Osteosarcomain Vivo and In Vitro

  Updated version

  http://cancerres.aacrjournals.org/content/55/6/1321

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

  [email protected] 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/55/6/1321To request permission to re-use all or part of this article, use this link

on May 23, 2020. © 1995 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from