ICANCER RESEARCH57. 5309-5319, December 1, 1997J

ABSTRACT

Whereas in advanced metastatic meduliary thyroid cancer (MTC), avariety of chemotherapeutic regimens have achieved only limited successclinically, more recently, radioimmunotherapy (RIT) with ‘311-labeledanti-carcinoembryonic antigen (CEA) monoclonal antibodies (MAbs) hasshown promising results. The aims of this study were to compare, in ananimal model, the therapeutic efficacy of lilT to clinically used “standard―chemotherapeutic regimens and to evaluate whether combinationstrategies of both modalities may be feasible and may help to improve

therapeutic results in this rather radioresistant tumor type.Nude mice, bearing s.c. xenografts of the human MTC cell line, TT,

were treated either with the 131I-labeledanti-CEA MAb, F023C5 IgG, orwere administered chemotherapeutic regimens that had shown promising

results in patients with metastatic MTC (doxorubicin and cisplatinum

monotherapy, combinations of both agents, and a 5-fluorouracil/dacarbazine/streptozotocin scheme). Control groups were left untreated or wereinjected with an irrelevant radiolabeled antibody at equitoxic dose levels.

The maximum tolerated dose (MTh) of each agent was determined.Combinations of chemotherapy and RIT were evaluated as well. Toxicityand tumor growth were monitored at weekly intervals.

From the chemotherapeutic agents and schemes tested, doxorubicinmonotherapy was the most effective; combination therapies did not resultin an increased antitumor efficacy, but they did result in more severetoxicity. At equitaxic doses, no significant difference was found betweenthe therapeutic efficacy ofdoxorubicin and that of RIT. Myelotoxicity wasdose limiting with radiolabeled MAbs (MTD, 600 @sCi),as well as withchemotherapeutic regimens containing alkylating agents (cisplatinum,dacarbazine, or streptozotocin). At its MTh (200 pg), doxorubicin causedonly mild myelotoxicity, and despite signs of cardiac toxicity, gastrointes

tinal side effects were dose limiting. Accordingly, bone marrow transplantation (BMT) enabled dose Intensification with RIT (MTh with BMT,1100 g&Ci), which led to further increased antitumor efficacy, whereasBMT was unable to increase the MTh of doxorubicin. Due to the complementality of toxic side effects but an anticipated synergism of antitumorefficacy, combinations of lilT with doxorubicin were tested. Administra

tions of 500 @.tCiof ‘31I-labeled anti-CEA and, 48 h later, 200 ,.ag of

doxorubicin (Le., 83 and 100% ofthe respective single-agent MTDS), werethe highest doses that did not result in an increased lethality; with bonemarrow support, 1000 @aCiof ‘311-labeledanti-CEA could be combinedwith 200 gagof doxorubicin (Le., 90 and 100% of the individual MTDs).Therapeutic results of this combined radloimmunochemotherapy were

superior to equitoxic monotherapy with either agent, and indication for

synergistic antitumor effects is given. At its respective MTh, radioimmunochemotherapy led to a 36% cure rate if it was given without bone

Received 6/23/97; accepted 10/9/97.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 to indicate this fact.

@ This work was supported in part by a research grant from Henning Berlin (Berlin,Germany) and by Deutsche Forschungsgemeinschaft Grant Be 1689/4-1.

2 To whom requests for reprints should be addressed, at Department of Nuclear

Medicine, Georg-August-University of Gouingen, Robert-Koch-Strasse 40, D-37075Gottingen, Germany. Phone: 49-551-39-8510; Fax: 49-551-39-8526.

marrow support and to a 85% permanent cure rate if it was given withbone marrow support.

The animal model, as presented in this study, seems to be useful for thepreclinical testing of therapeutic agents for the systemic treatment of

MTC. At equitoxic doses, RIT with radiolabeled anti-CEA antibodiesseems to be equally as effective as chemotherapy with doxorubicin. Combination of kIT and doxorubicin chemotherapy seems to have synergistictherapeutic efficacy, which may be due to a radiosensitizing effect ofdoxorubicin.

INTRODUCTION

MTC3 is a comparably rare neoplasm. It accounts for less than 10%of the thyroid cancers that are diagnosed every year. Its annualincidence comprises only approximately 3500 new cases in Westernpopulations ( 1, 2). However, recent systematic screening studies suggested that there may be a higher incidence of clinically nonapparentforms of MTC (3).

As long as MTC is confined to the thyroid gland itself, cures can beachieved by total thyroidectomy and cervical lymph node dissection(1, 2, 4—7). Frequently, however, postsurgically elevated plasmacalcitonin and/or CEA levels indicate persisting metastatic disease ( 1,2, 8). Mediastinal lymph nodes were identified as the most frequentsite of disease in this setting (9, 10). The prognosis of patients withmetastatic disease confined to mediastinal lymph nodes is mostlygood, whereas the prognosis of patients with distant metastases (e.g.,in the liver, lung, or bone) is poor ( 1, 2, 10); their 5-year survival isbelow 30%, and these patients are left with very few therapeuticchoices (1, 2, 4, 6—8). External beam radiation is useful for control

ling local disease (7), not for widespread metastatic involvement, even

more so because MTC is generally regarded as a comparably radioresistant tumor type (1 1). A variety of chemotherapeutic regimens havebeen tested, but despite encouraging single-case reports, most of themwere found to be of little value in larger controlled series withreported objective response rates that were generally below 30% ( I,2). Thus, novel therapeutic strategies seem to be warranted.

Recently, Juweid el a!. (12, 13) have shown promising clinicaltherapeutic results using I31I-labeled anti-CEA antibodies for thetreatment of advanced metastatic MTC. Although it is radioresistant,MTC seems to be a good candidate for a therapeutic approach withradiolabeled anti-CEA antibodies because we have shown that MTChas a significantly higher CEA expression than do other cancer types(14). Thus, significantly higher antibody uptake values and, consequently, tumor doses are found in this cancer type than in any otherCEA-expressing tumor (14), independent of the patients' actual CEAlevels in blood (10, 12—14).Accordingly, initial clinical results with

3 The abbreviations used are: MTC, medullary thyroid cancer; CEA, carcinoembryonic

antigen; RIT, radioimmunotherapy; MTD, maximum tolerated dose; MAb, monoclonalantibody; HPLC. high-performance liquid chromatography; 5-FU, 5-fluorouracil; BMT,bone marrow transplantation; CK, creatine kinase; %lD, percentage of injected dose; p.i.,postinjection: Gb, gastrointestinal; RIO', radioimmunochemotherapy.

5309

Improved Treatment of Medullary Thyroid Cancer in a Nude Mouse Model byCombined Radioimmunochemotherapy: Doxorubicin Potentiates the

Therapeutic Efficacy of Radiolabeled Antibodies in aRadioresistant Tumor Type'

Thomas M. Behr,2 Erik Wuist, Sven Radetzky, Rosalyn D. Blumenthal, Robert M. Dunn, Stefan Gratz,Margret Rave-Frank, Heinz Schmidberger, Friedheim Raue, and Wolfgang BeckerDepartments ofNuclear Medicine (7'. M. B., E. W., S. R., S. G., W. B.J and Radiation Oncology (M. R-F.. H. Si. Georg-August-Unis'ersitv. D-37075 Grittingen. Germany: GardenState Cancer Center, Belleville, New Jersey 07109 (R. D. B., R. M. DI: and Department of Endocrinology. Ruprecht-Karl-Unit'ersizv. Heidelberg, Germany (F. R.J

on April 5, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

IMPROVED TREATMENT OF MTC BY RICT

RIT of MTC were encouraging; despite the fact that most patientswere treated below the MTD, the therapeutic results were comparableto those of most published chemotherapy series (12, 13).

Despite these encouraging clinical data, a more systematic preclinical investigation of systemic approaches for treating advanced MTC

seemed to be warranted. Therefore, the aim of this study was toestablish a suitable animal model of human MTC in nude mice todetermine and compare the dose-limiting organ toxicity, the resultingMTDs, and the antitumor efficacy of RIT to those of different chemotherapeutic regimens that had shown some effectiveness in clinical

studies. Finally, we wanted to investigate whether combination strutegies of RIT and chemotherapy may be feasible and may help toimprove therapeutic results, especially because several of the chemotherapeutic agents used are known as radiosensitizers (1).

MATERIALS AND METHODS

Human MTC Model in Nude Mice. The humanMTCcell line, iT, wasobtained from the American Type Culture Collection (Rockville, MD). This

cell line was obtained by a needle biopsy of a medullary thyroid primary tumorfrom a 77-year old Caucasian female, and it produces high levels of CEA andcalcitonin (15). Cells were cultured at 37°Cin Ham's F-l2K medium (ICNBiomedical, Eschwege, Germany) supplemented with 10% FCS (Sigma Chemie, Deisenhofen, Germany). Genetic analysis showed that TI' is a heterozygous carrier of a mutation in codon 634 of the RET proto-oncogene (TGC —@TOG, corresponding to an amino acid exchange Cys —@Trp).4

Approximately l07@I08 cells were inoculated s.c. into female nude mice,

weighing 19—23g and 4—5weeks old (Charles River, Sulzfeld, Germany). Thetumors were serially propagated by preparing a mince through a 40-meshscreen and rinsing with Ham's F-l2K medium supplemented with 20% FCS toyield a 20% cell suspension. Two hundred fifty pi of this suspension wereinjected s.c. After approximately 3—4weeks, over 90% of animals had devel

oped tumors of approximately 100—200mg, which was the size used in thisstudy unless otherwise indicated. Supplementation of the medium with serumwas crucial for obtaining reproducible tumor take and growth rates. The humancolon carcinoma cell line, GW-39 (16, 17), was used for comparing radiationdose-antitumor efficacy relationships and radiosensitivity.

Antibodies, Radionuclides, and Radiolabeling. The murine anti-CEAMAt,, F023C5, was obtained from Sorin Biomedica (Saluggia, Italy). Itsdevelopment and clinical use with ‘@ ‘I,‘‘‘In,or @“Fclabel has been describedpreviously (18, 19). F023C5 is an IgGI isotype that is directed against a classIII epitope of CEA (18, 19), according to the classification of Primus et a!.(20). It has an affinity of 0.5 X 108 liters/mol to its epitope. The antibody wasstored frozen at —20°Cat a concentration of 10 mg/ml in PBS buffer at a pH

of 7.2. The anti-CD3 antibody, OKT3, was used as nonspecific control, whichwas obtained from CILAG (Sulzbachll'aunus, Germany).

Radioiodination with Na'31! in 0.1 M NaOH (New England Nuclear/DuPont, Bad Homburg, Germany) was performed using the iodogen method,essentially as described previously (21, 22). Briefly, the antibody in PBS bufferwas transferred into a iodogen-coated glass vial (500 @gof iodogen coating theinner surface of a 10-mI vial) with a magnetic stirbar placed inside. Sodiumphosphate buffer (0.5 M;pH 7.4) was added in a volume that was twice as highas the volume of the radioiodine to be used. The specific activity used was35—40 mCi/mg of antibody protein. The vial was placed on a magnetic stirrer,

and the activity was added in 1.5 ml of 0.04 M sodium phosphate (pH 7.4).After a stirring time of 5—10mm, Dowex lX8-l00 anion exchange resin (CFform; Sigma) was added, and the incubation time was prolonged for another 5mm. Subsequently, the radioiodinated antibody was filtered through a sterileMillex-GV filter (pore size, 0.22 @m;Millipore, Molsheim, France). Thequality of each preparation was tested by instant thin-layer chromatographyand size-exclusion HPLC. The amount of unbound radioiodine was less than2% in each preparation. The immunoreactivity was 90% in all cases, astested by a complexation assay by incubation with a molar excess of CEA,followed by size-exclusion HPLC analysis.

Biodistribution Studies. Tumor-bearing animals were injected i.v. into thetail vein with the ‘311-radiolabeledantibody. The mice were sacrificed at 1, 4,24, 72, 168, and 336 h. They were bled by retro-orbital puncture. After cervical

dislocation, the animals were dissected. The amounts of activity in the tumor

and tissues (liver, spleen, kidney, lung, intestine, blood, and bone) weredetermined by gamma scintillation counting using an injection standard to

account for physical decay, as described previously (17, 23). The number of

animals used for each study was typically five animals per group at each timepoint, and the tumor sizes were in the same range as those used for the actualtherapy experiments, unless otherwise indicated.

Radiation Dosimetry. The biodistribution data were used to generatetime-activity curves (17, 24), which were fit to an exponential function andintegrated over time to obtain cumulated activities (24). If the regressioncoefficient (r) for this exponential fit was worse than —0.90(i.e., rJ < 0.90),the integral was computed trapezoidally instead (17, 24). Mostly, the cumulated activities of normal organs were calculated from the exponential model,whereas for tumors, the trapezoidal model was used (17, 24). Radiationabsorbed doses for 13‘Iwere calculated from the cumulated activities according to the model developed by Siegel and coworkers (17, 25), with modifica

tions.5 For all tissues except bone, the assumption of uniform distribution ofactivity in a spherical mass was maintained, and doses were calculated using

S-factors generated for the respective radionuclide in spheres of 0.1—1g(25,26).

Chemotherapeutic Agents. Doxorubicin hydrochloride (Adriamycin) wasobtained from Pharmacia (Milan, Italy) as a solution of 2 mg/mI in 0.9% saline.Cisplatinum was obtained as a solution of 0.5 mg/mI in 0.9% NaCI from HexalPharma (Holzkirchen, Germany). 5-Ri, dacarbazine (5-[3,3-dimethyl-1-triazenyl)imidazole4-carboxamide), and streptozotocin (N-[methylnitrosocarbamoyl]-n-glucosamine) were purchased from Sigma. Solutions were prepared

in sterile saline to give final concentrations of 2.5 mg/mI each.Experimental RIT and Chemotherapy. Tumor sizes were determined by

caliper measurement in three dimensions immediately before therapy and atweekly intervals thereafter. Tumor volumes (V) were estimated by multiplying

the product of the three perpendicular diameters (d,, d2, and d3) by 0.5(V = d, x d2 X d3 x 0.5), assuming an elliptical geometry. Tumors were

either left untreated (controls), injected with a single dose of radiolabeledantibody, or given chemotherapy, with the activities and dose schedulesindicated below. Eight to 20 animals were studied in each treatment group.

Body weight was recorded weekly, and survival was monitored. The MTDwas defined as the highest possible activity under the respective conditions thatdid not result in any animal deaths, with the next highest dose level resultingin at least 10% of the animals dying. Animals were observed until their death

or a until they had lost >30% of their original weight, at which time they wereremoved from the group and sacrificed.

BMT for Dose Intensification. Bone marrowwas harvestedby steriletechnique from untreated donor nude mice of the same strain. The marrowcavities of both mouse femurs were rinsed with 0.9% sterile saline, as de

scribed previously (27). An inoculum of approximately l0@bone marrow cellswas injected i.v. via the tail vein 120 h after radioantibody or chemotherapyinjection (17, 27).

Determination of Blood Counts and Liver, Renal, and Cardiac Toxicities. Total and differential WBC and platelet counts were determined on theday of therapy and at weekly intervals thereafter. Heparinized specimens (75s.d) were collected by retro-orbital bleeding. The samples were counted on a

Technicon H3 Auto-Analyzer (Bayer-Diagnostik, Munich, Germany). Bloodurea nitrogen and creatinine as indicators of renal function, CK as a parameterof cardiac toxicity, and glutamate oxaloactetate transaminase, and alkalinephosphatase as liver function parameters were determined at the same timepoints as the blood counts. All these parameters were assayed in a fullyautomated system (17). Means ±SDs were calculated for each group.

Analysis of the Antitumor Efficacy. The correlationbetween calculatedmean tumor doses and the extent of induced growth retardation as a parameterof antitumor efficacy was analyzed. Because s.c. growing tumors usually havetumor volume doubling times of

IMPROVED TREATMENT OF MTC BY RICF

delivered to the tumor within this time frame. Therefore, mean tumor doseswere correlated to the mean time needed for the quadruplication of tumorvolumes (i.e., two doubling times).

The Loewe additivity model as described by Greco et a!. (28) was used todetermine whether the combination of RIT and chemotherapy had additive orsynergistic antitumor effects (28, 29). In accordance with the modeling ofKievit et a!. (29), the mean relative tumor volumes at 2, 4, and 6 weeks aftertherapy were chosen as the end points. The dose-response relationships of bothMT and chemotherapy were assumed to be exponential and were plotted aslogarithmic relative tumor volume versus activities of radiolabeled antibody orchemotherapeutic doses (29, 30). The intercepts of the dose-response curves of

the single-treatment modalities with the mean relative volumes of tumors thatwere treated with the combination therapy were determined. These interceptswere assumed to represent the doses of radiolabeled antibody or chemotherapyalone that were expected to be isoeffective, as compared to the combinationtherapeutic doses (29). The effect of the combination treatment was considered to be synergistic (if dR,TDRIT â€+̃ dchem@Dchemoâ€̃ 1), with dR,.,. and dchemo being the activities of radiolabeled

antibody or chemotherapeutic doses that were actually given in the combinaHon and Dv,.,.and Dc@mobeing the calculated activities and doses that wereexpected to be equally effective when given alone (28, 29).

Statistical Analysis. Differences in the efficacy between the treatmentmodalities were statistically analyzed with the Student's t test for unpaireddata, as described previously (17).

RESULTS

Factors Influencing the Biodistribution of the RadiolabeledAnti-CEA Antibody, F023C5. Fig. 1 shows the biodistribution of‘311-labeledF023C5 in the tumor, liver, and blood in s.c. U xenograft-bearing nude mice. As compared to other murine IgGls MAbs(e.g., see Refs. 17 and 24), F023C5 showed a fairly rapid bloodclearance, with a t½aof —1 h and a t½jsof —50 h, which wasaccompanied by a rapid and comparably high liver uptake (e.g.,8.0 ±1.8 %ID/g at 4 h p.i.; Fig. 1). Maximum tumor uptake wasreached at 24 h p.i. and was strongly dependent upon the tumor size(Fig. 2A), with uptake values ranging from 0.2 to 12.4 %ID/g intumors between 0.1 and 2.0 g, with a mean of 8.8 ±1.4 %ID/g at 24 hp.i. for 100-200-mg tumors.

Because of the unexpectedly rapid blood clearance and enhancedliver uptake, we investigated, in more detail, which factors mayinfluence the pharmacokinetic behavior of F023C5 IgG. Fig. 2, B andC, shows that the blood clearance and the liver uptake seem to dependupon the mass of tumor tissue present. With increasing tumor size,both liver uptake and blood clearance increased. Furthermore, external scintigraphy showed uptake of metabolically liberated iodine inthe thyroid (compare Fig. 1B), which could be prevented by theaddition of potassium iodide to the animals' drinking water supply(data not shown). Because the absolute tumor uptake (i.e., the %IDper total tumor) did not differ significantly between animals withsmall tumors and those with larger ones (ranging between 0.4 and 2.0%ID/tumor), these differences in antibody clearance cannot be explained by differences in the amount taken up by the tumor. Incontrast, an increasing liver uptake due to hepatic clearance of immune complexes between the injected antibody and circulating antigen was described previously for CEA or other antigens that are shedby the tumor into the circulation (22, 31—33).Although no quantitative tumor marker determinations in mouse blood or HPLC analyseson the relationship between tumor size and complexation have beencarried out in this study, such an hepatic clearance of immune complexes between antibody and circulating CEA appears as the mostlikely explanation for the observed pharmacokinetic differences.

Table 1 summarizes the radiation dosimetry and tumor:nontumordose ratios, based on the biodistribution data for animals bearing

A01

0

15

0.

B

F―2d―@51311I@1 @?@# .-.g : tumor

.@o- blood

—.-- liver

100

time p.i. (h)

thy

ii

.@,

. - @:@

...@

•1 @I @tS@ -

Fig. 1. Biodistribution of the ‘@‘I-labeledanti-CEA antibody. F023C5, in nude micebearing s.c. xenografts of the human MTC cell line, iT. A, biodistribution in the tumor,liver, and blood. B, extemal scintigraphy of a T1'-xenografted nude mouse at 72 h p.i.(Picker Prism 2000 gamma camera with high-energy, parallel-hole collimators; 100kilocounts), showing antibody uptake in a large s.c. ‘Ii'tumor (arrow), as well as acomparably high liver (Ii) accretion, probably due to hepatic uptake of immune complexesbetween the injected MAb and circulating CEA. Thy, uptake of metabolically liberatediodine in the thyroid.

100—200-mgtumors. At tumor doses of approximately 60 Gy/mCiand blood doses of 25 Gy/mCi, the comparably high liver uptake ledto liver doses of 29 Gy/mCi, whereas all other organ doses were atleast 3-fold lower.

Experimental Therapy of s.c. TT Xenografts: Therapeutic Efficacy and Toxicity of RIT. Animals bearing U s.c. xenografts(8—20animals per treatment group) were treated with ‘311-labeledF023C5 to determine its MTh, induced treatment-related toxicity, andantitumor efficacy. To establish the MTD of ‘@‘I-labeledF023C5,varying amounts of activity were injected, starting at 250 p.Ci andincreasing in 10—20%steps. Acute treatment-related death was defined as occurring within 6 weeks after radioantibody injection.

5311

on April 5, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

TissueDose (cGy/mCi)Tumor:NontumorratioDoseat MTIY'

(Gy)Dose

at MTDwith BMr

(Gy)Tumor593135.665.2Liver28512.117.131.4Spleen9836.05.910.8Kidney13114.57.914.4Lung9236.45.510.1Intestine8197.24.99.0Bone39914.92.44.4Blood25672.315.428.2

IMPROVED TREATMENT OF MTC BY RICT

myelosuppression occurred (Fig. 3), and a further 10% increase in theadministered activity resulted in 10—20%death at 3—4weeks afterradioantibody administration. Nadirs of WBC and platelet countswere reached at 2 weeks after therapy, and full recovery was observedat 4—5weeks (platelets recovered sooner than leukocytes). No othersigns of second-organ toxicity were observed; glutamate oxaloactetatetransaminase and alkaline phosphatase, as indicators of liver function,and blood urea nitrogen and creatinine, as parameters of kidneyfunction, did not show any significant changes, which is in accordanceto normal organ doses well below 20 Gy (compare Table 1).

To further examine the role of the bone marrow as dose-limitingorgan, BMT was given on day 5 after radioantibody injection. Thisallowed for an increase in the maximum tolerated activity by approximately 80%, to 1100 pCi of ‘311-labeledF023C5. No signs ofsecond-organ toxicity other than severe myelosuppression were seenat this dose level either, despite radiation doses to the liver of >30 Gy(compare Table 1).

—.— controls

-..... 1311-anti-CEA

--A-. cisplatinum

-@- 5-FU/dacarb./strept

..4.. doxorubicin

—0-@ controls

-.o@. 1311-anti-CEA

--A-S cisplatinum

-@- 5-FU/dacarb./strept

.@o.. doxorubicin

.$ll'

•. S

SS

S@@

S

SS

0.0 0.5 1.0 1.5 2.0

tumor weight (g)

A@ 10.0

0-@.@

B01a

C0

C0

•000

.0

C2001

@15

,@ 100.

I-05>

-.-- no tumor

—0- 0.12 ±0.09 9

-@- 1.45±0.53g

time p.1. (h)

S•

S S@ 5 @.—.----

S@@@@ S

__@-@--——--;__@ S@

5• S

-1 0 1 2 3 4 5 6 7 8 9

weeks posttherapy

_0.0 0.5@ 1.5 2.0

tumor weight (g)

Fig. 2. Factors influencing the biodistribution of the anti-CEA MAb, F023C5. A,relationship between tumor uptake (at 24 h p.i.) and tumor weight. B, dependence of theblood clearance upon the tumor mass. C, liver uptake at 4 h p.i. in relation to the tumormass. 1234567

weeks post therapy

Table 1 Radiation dosimetry of 1311-labeledF023C5 igG in s.c. 77' xenograft-bearingnude mice (with tumor sizes of 100—200mg)

Li_J_______W_S •animalweight0 creatine kinase

aTheMTDisdefinedasthehighestpossibleactivityundertherespectiveconditionsthat did not result in any animal deaths, with the next highest dose level resulting in at least10%of the animals dying. The MTDs of ‘3'I-labeledF023C5 IgG were identified as 600@.&Ci,without artificial support, and as 1100 pCi, with bone marrow stem cell support.

a 50 100 150 200 250doxorubicin (pg)

Fig. 3. Toxicity of Rif with ‘31I-labeledF023C5, as compared to chemotherapy withdoxorubicin, cispbatinum, or the 5-FU/dacarbazine/streptozotocine scheme tested (withexception of C, all values shown are at their respective M'IDs; for details on dosage andscheduling, see text). A, WBC counts. B, platelet counts. C, relationship between doxorubicin dosage and the resulting weight loss or serum CK elevation. Left Y axis (filledsymbols and boldface lines), percentage of weight loss (negative values representing anactual weight increase). Right Y axis (open symbols), CK elevation as n-fold overpretreatment levels.

Whereas at a 250-pCi level of ‘311-labeled F023C5 (corresponding to

a blood dose of approximately 6.4 Gy), no significant leukopenia andthrombocytopenia were observed, the MTh was reached at 600 p.Ci(corresponding to 15 Gy to the blood). At this dose level, severe

5312

A

,@C

0

I

B

,@C

0U

2150.

C50

@.40

@30

@ 20

@io

E -io

-2011

on April 5, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

@..@hJLI'.

IMPROVED TREATMENT OF Mit BY RICT

sions (“cures,―defined as complete disappearance of s.c. tumorswithout signs of regrowth over the whole observation period).

Experimental Therapy of s.c. TT Xenografts: Therapeutic Elficacy and Toxicity of Various Chemotherapy Regimens. Different chemotherapeutic regimens that were the most frequently used and

that had shown some therapeutic results in clinical trials were tested(1, 2, 34—36), namely, doxorubicin, cisplatinum, combinations ofboth, and a 5-FU/dacarbazine/streptozotocin regimen that was recently introduced as promising alternative to doxorubicin therapy bySchlumberger et a!. (35). As in the RIT dose-finding studies, theMTDs of the different agents were determined by administeringvarying amounts of the respective drug in 10—20%increment steps.

Doxorubicin was the first chemotherapeutic drug to show antitumorefficacy in different forms of thyroid cancer, including MTC (1, 2, 34,36). When given as single i.v. dose in our nude mouse model, 200 @gwas shown to be the MTh. Animals treated in this manner lostapproximately 20—30%of their original weight within 2 weeks, andthey mostly experienced severe diarrhea as the indicator of GI toxicity(Fig. 3C). In approximately one-third of animals, transiently rising CKvalues up to 3 times the pretreatment levels were found, which may beindicative of cardiac toxicity. However, CK values returned to baseline within 1 week, and no clear dose-effect correlations between theamount of doxorubicin administered per animal and the extent of CKelevation were found (Fig. 3C). Myelotoxicity caused by doxorubicinwas comparably mild (leukopenia > thrombocytopenia; Fig. 3, A andB), and no evidence of hepatic or renal toxicity was observed.

When mice were treated at the MTD (200 @.tg/mouse),doxorubicinled to a significant (P < 0.02) growth retardation, as compared tountreated controls, which was similar to the growth delay seen withRIT (compare Figs. 4 and 5). However, the variability in growth delaywas significantly (P < 0.05) greater than that observed with RIT(compare Figs. 4 and 5). Similar to lilT treatment, 2 of 10 animalstreated with doxorubicin at its MTh experienced enduring completeremissions.

In contrast to the rather mild myelotoxicity observed with doxorubicin, the red marrow was the dose-limiting organ with cisplatinum.The single-administration MTh was identified as 125 @.tg,which ledto a similarly severe leukopenia as that seen with RIT, whereasthrombocytopenia was less pronounced (Fig. 3, A and B). However,with respect to its antitumor efficacy, no significant differences wereobserved to untreated controls (Fig. 5). Combinations between doxorubicin and cisplatinum were tested as well. Antitumor effects werenot superior to doxorubicin monotherapy, but myelotoxicity was morepronounced (data not shown).

Because promising antitumor effects against metastatic MTC havebeen reported in a clinical trial with an alternating 5-FU/dacarbazine5-FU/streptozotocine scheme (with objective response rates of —15%;Ref. 35), we tested the therapeutic efficacy of such a regimen. Tomimic the clinical regimen reported by Schlumberger et a!. (35), five

daily injections of 5-FU and dacarbazine were followed after a 1-weektherapy-free interval by five daily injections of 5-Ri and streptozotocine. As was seen with cisplatinum, myelotoxicity was severe anddose limiting. In the scheme tested, the MTD was reached at five dailyinjections of 0.5 mg of 5-Hi and 100 @gof dacarbazine, followed byfive daily injections of 0.5 mg of 5-FU and 100 .tg of streptozotocine7 days later. The antitumor efficacy of this regimen was only borderline significant (P 0.06), as compared to untreated controls (Fig. 5).

BMT was unable to increase the MTDs of doxorubicin, which is inaccordance to the low myelotoxicity observed. Due to their disappointing antitumor efficacy at conventional dosing, no studies on theeffects of BMT on the other chemotherapeutic drugs were undertaken.

15

IE

0

15

EU

IE‘a

0

15

@ 10a

15@

0

a

E.a

0

controls

—5—1311-anti-CEA

-.- 131l-anti-CEA + BMT

.-e.. irrelevant 1311-lgG

10

weeks post therapy

Fig. 4 shows the tumor growth curves of animals treated at therespective MTDs without and with bone marrow support, as comparedto untreated controls or animals given equitoxic doses of ‘311-labeledirrelevant IgG (‘311-OKT3).In Fig. 4, lightface lines represent therespective tumor volumes of individual animals, whereas boldfacelines represent the means of the respective treatment groups (errorbars represent SDs). Whereas untreated controls grew with meantumor volume doubling times between 1 and 2 weeks, tumor growthwas retarded significanfly (P < 0.01) by the radiolabeled anti-CEAantibody. In animals treated at its MTD (600 @aCiof ‘311-labeledF023C5 IgO), the tumor volumes 3 weeks after therapy were64 ±29%of theinitial volumes,ascomparedto 491±395%in theuntreated controls. In contrast, unlabeled F023C5 protein (in doses upto 500 @.tg,data not shown) or equitoxic doses of ‘311-labeledOKT3as irrelevant antibody did not have any significant effect on tumorgrowth (Fig. 4). When administering ‘311-labeledF023C5 at its MTDwith BMT (1100 ,.aCi), antitumor effects were significantly(P < 0.01) better than they were at the conventional MTD (600 @aCi).Two of 12 treated animals even experienced durable complete remis

Fig. 4. Comparison of the therapeutic efficacy of the ‘3'I-labeledspecific antibody,F023C5, at its MTD without (600 pCi) and with (1 100 @&Ci)bone marrow support, ascomparedto untreatedcontrolsor animalsequitoxicallytreatedwithirrelevantradioantibody (‘311-OKT3).lightface lines, respective tumor volumes of individual animals.Boldface lines, means of respective treatment groups; bars, SD.

5313

on April 5, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

IMPROVED TREATMENT OF MTC BY RICT

to two deaths in a group of 20 animals. The animals treated at thecombination MTD (500 @Ciand 200 p.g), lost 20—30% of theiroriginal weight within 2—3weeks and showed signs of GI toxicitysimilar to those seen in animals that had been treated with doxorubicinonly. Accordingly, the animals had regained their original weight in3—4weeks after therapy. With this combination therapy, myelotoxicity was more prolonged than it was with RIT only; the nadir,

however, was similar to that observed with RIT alone. The blood cellcounts recovered within 6—7weeks. As observed with single-agenttherapy, no changes of renal or hepatic function parameters werefound with this combined RICT regimen.

With BMT, given 120 h after radioantibody injection, 1000 @tCiof‘31I-anti-CEA(i.e., 90% of the single-agent MTh with BMT) couldbe combined with 200 @gof doxorubicin, whereas the combination of1 100 @Ciwith 200 @gof doxorubicin resulted in the deaths of 2 of 11

animals. The BMT of the animals treated at the MTD was not

significantly different from the myelotoxicity observed with RITalone. Additionally, the animals in the BMT group developed atransient skin rash, resembling radiation dermatitis, which resolvedwithin 2—3weeks after therapy. Again, some animals (8 of 20) hadtransient CK elevations at 1 week posttherapy, and values returned tobaseline within 1 week.

Fig. 6 shows the therapeutic results of this combined RICT. Combination of 500 @Ciof ‘311-anti-CEAwith 200 @gof Adriamycin, the“conventional―MTD, led to significantly (P < 0.01) better tumorgrowth retardation than did either treatment modality alone, with 4 of11 animals remaining in complete remission over a >30-week observation period. Animals given 1000 @Ciand 200 p.g with bone marrowsupport showed best therapeutic results, with a 85% (17 of 20) curerate.

Improved Therapeutic Results Are Due to Doxorubicin's Aclion as Radiosensitizer: Dosimetric Considerations. Fig. 7A showsthe relationship between calculated tumor doses and the resultingantitumor effectiveness. Because tumor growth retardation was themajor observed effect in all treated groups, the correlation betweenmean tumor doses and the extent of induced growth retardation wasanalyzed. Fig. 7A shows the correlation between tumor doses and themean time needed for the quadruplication of tumor volume (i.e. , twodoubling times) as compared to untreated controls. Below a tumordose of approximately 30—40Gy, no significant tumor growth retardation was noticeable with ‘311-labeledantibodies. Above this threshold, tumor growth was retarded in a dose-dependent fashion, up to10-fold, at approximately 65 Gy (i.e., the tumor dose at the MTh withBMT; compare Table 1). This behavior identifies ‘VTas a relativelyradioresistant cell line. We have shown earlier (17, 39) that in othersolid tumor lines, such as in colorectal cancer, significant growthinhibition is achievable at much lower radiation doses by radiolabeledantibodies. Therefore, Fig. 7A shows, for comparison, the relationshipbetween radiation dose and growth inhibition in the GW-39 humancolon cancer cell line. With this cell line, significant effects are seenalready at doses above approximately 15 Gy (Fig. 7A). Interestingly,when RIT is combined with doxorubicin in TI' tumors 48 h later, the“shoulder―of the resulting tumor response curve is shifted to lowerdoses, resembling the curve of the more radiosensitive colorectal lineGW-39, whereas doxorubicin alone (corresponding to a radiation doseof 0 Gy to the tumor) merely led to an approximately 2-fold growthdelay in TI' tumors (Fig. 7A).

Fig. 7B shows the cumulated doses (filled symbols and solid lines),as well as dose rates (open symbols and doued lines), in tumors andblood in the time course after radioantibody injection. The figureshows that, at 48 h p.i., the very time of Adriamycin administration,the dose rate to the blood and, thus, the red marrow, has fallen to lessthan 20% of its initial value, whereas the dose rate in the tumors is still

5314

15-

@ 10@

@ 5-

E‘a.

0-

15

E@2.@ 10@

!@@0 @4if@

—5— controls

—0— 5.FU/dacarb./strept.

—6-- cisplatinum

- @- doxorubicin

15CIEU

@1E‘a

0

15

@1Ea4'

0

.-.

@l 1@0

weeks post therapy

Fig. 5. Comparison of the therapeutic efficacy of several clinically common chemotherapeutic drugs and regimens in MTC (doxorubicin, cisplatinum, and 5-FU/dacarbazine/streptozotocine) at their respective MTDs. as compared to untreated controls (for details,see text). Lightface lines, respective tumor volumes of individual animals. Boldface lines,means of respective treatment groups; bars, SD.

Toxicity and Antitumor Efficacy of a Combination Strategy ofRIT and Chemotherapy with Doxorubicin. Because doxorubicinand RIT with ‘311-labeled anti-CEA antibodies were identified as themost effective treatment modalities tested in this study and becausetheir toxic side effects seemed to affect different organs (0! tract andpossibly heart versus bone marrow), combinations of both agents may

be feasible and may enhance the therapeutic efficacy; this is evenmore true because doxorubicin is known as a potent radiosensitizer (1,37, 38). Therefore, a combination regimen was tested, with the radiolabeled antibody administered first, followed by doxorubicin 48 hlater. This schedule was chosen because, at 48 h after radioantibodyadministration, over 90% of its activity has been cleared from blood,

whereas the tumor uptake is still close to its maximum (compare Fig.1A).

Again, MTD finding studies were undertaken. It was possible toadminister, without increasing the lethality, 500 p@Ciof ‘311-anti-CEAMAb (i.e., 83% of the single-agent MTD) and 200 @gof doxorubicin(i.e., the full MTD dose) 48 h apart, whereas 600 pCi and 200 p.g led

on April 5, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

,i/@/

@—‘@-:@-TT1Trru

. .@ . !i@:rrr

IMPROVED TREATMENT OF MTC BY RICT

15

EU

@ 1;

- controls

—— 131l-anti-CEA+doxorubicin

- @- 1311-anti-CEA+doxorubicin+BMT

Fig. 6. Therapeutic efficacy of a combination therapy of 1311-IabeledF023C5 and doxorubicin, given 48 h apart, at the MTD without (500 @Ciof‘3'I-labeled F023C5 and 200 @&gof doxorubicin) and with (1000 @sCiof‘3'1-labeledF023C5 and 200 @gof doxorubicin) BMT, as compared tountreated controls. lightface lines, respective tumor volumes of individualanimals. Boldface lines, means of respective treatment groups; bars, SD.

close to its maximum. This may well explain why no major effect wasobserved on the radiosensitivity of the red marrow, whereas effects inthe tumor were marked.

Finally, we investigated whether the improved antitumor efficacyof the combination therapy is due to a real synergistic effect or just an

a i'o @o 30weeks post therapy

additive phenomenon. The dose-response curves of both ‘311-labeledF023C5 and doxorubicin monotherapy showed an exponential pattern(Fig. 8, open symbols with lightface lines; linear fit in semilogarithmic scale, with r 0.9 in all cases). The mean relative tumor volumesat 2, 4, and 6 weeks after combination therapy were plotted as

A20

SEso@ 15

@ 10

.!E 5

B

0 U TI: ‘311-anti-CEA0 TI: l3l@4@j@

. GW-39:‘31l@fl@J@@

Fig. 7. Dosimetric considerations. A, correlation between calculated tumordoses and the resulting antitumor effectiveness (expressed as the time to tumorvolumequadruplicationrelativeto untreatedcontrols)for TI' xenograftswithout and with doxorubicin, as compared to a colorectal cancer cell line(GW-39) as a more radiosensitive tumor type (16, 17).B, cumulated doses (•and •),as well as dose rates (and 0), in tumors and blood in the time courseafter radioantibody injection. Arrow, time of doxorubicin injection chosen inthis study (48 h p.i.).

0 20 40 60 80

tumor dose (Gy)

C,@30

0@0I20E 10aU

0

o@ ióo i@otime p.1. (h)

5315

on April 5, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

.

0 200 400 600 800. . Dchsmo

doxorubicin (pg)

Fig. 8. Dose-response curves of the ‘311-labeled anti-CEA antibody, F023C5 (A), and

doxorubicin (B) in iT xenografts on days 28 (El. 42 (a), and 56 (0). Horizontal lines,tumor volumes on days 28 (—), 42 (----), and 56 (. ) after combination therapy with500 @.sCiof F023C5 and 200 @gof doxorubicin, 48 h apart (for details, see text; Refs. 28and 29).

horizontal lines in Fig. 8. The intercepts of these lines with thedose-response curves of the single-treatment modalities were assumedto represent the hypothetical doses of the respective monotherapies(DRIT and Dchemo) that were expected to be as effective as thecombination (28, 29). Table 2 shows the resulting intercept valuestogether with the numbers resulting from the use of the formuladR,TDR,T + dchem,@Dchemo@ . We conclude from the equation values, ranging from 0.60 to 0.69, which are well below 1.0, that RIT and

doxorubicin have a synergistic therapeutic effect, which would be inaccordance with doxorubicin's mode of action as radiosensitizer.

DISCUSSION

Most systemic treatment approaches have been shown to be of littlevalue in aggressively metastatic MTC (1, 2). Best results were reported with doxorubicin (Adriamycin) monotherapy, but objective

response rates only range from 20 to 30% (1, 2, 34, 36). Recently,Juweid et a!. ( 12, 13) demonstrated encouraging therapeutic resultswith ‘311-labeledanti-CEA antibodies. CEA appears to be an excellenttarget molecule for a radioimmunotherapeutic approach because, un

Table 2 C'haracterization of the therapeutic efficacy of the combination treatment of1311.labeled anti-CEA antibodies anddoxorubicin@'dRIT

and dchemo are the actual doses of ‘@ ‘I-labeledF023C5 and doxorubicin giveninthecombination, whereas DRITand Dche@@.@@ the calculated doses of radioantibodyandchemotherapy

given alone that were expected to be isoeffective as compared totheircombination(according to Refs. 28 and29).Days

@ dchemo DRIT Dchemo dR,TDRJTposttherapy (MCi) (pg) (pCi) (pg) +dchem,,Dchemo28

500 200 1442 7850.6042

500 200 1305 6540.6956500 200 1367 6950.65a

Compare Fig.8.

IMPROVED TREATMENT OF MTC BY RICT

like calcitonin, which is a mainly intracellular antigen (2, 40), CEA isexpressed on the cell surface of >90% of MTCs (2) and is, thus, more

easily accessible for MAbs (40). Additionally, several studies, in vitro(41) as well as in vivo (10, 42), suggested that the amount of CEA

expression correlates to the degree of differentiation (the more ag

gressive and dedifferentiated the tumor, the higher the CEA expression). No significant correlation was found between the patients' CEA

levels in blood and the resulting tumor doses or the therapeutic

efficacy of radiolabeled anti-CEA antibodies ( 10, 12—14,22), indicating that MTC may well express high levels of CEA on the cell surfaceand may be well suited for anti-CEA RIT, although the cell surfacebound CEA may not be shed in larger amounts into the circulation.

Although the preliminary therapeutic results of Juweid et a!. (12,13) were encouraging and the observed tumor doses were significantly higher than in any other cancer type (14), they emphasized that,due to the general radioresistance of MTC, potentially myeloablativedosing may be necessary for improving therapeutic results (12—14).Although MAb uptake values of up to 0.9 %ID/g were observed (14),corresponding to tumor doses between 4.2 and 174.0 cGy/mCi (14,43), dose-limiting myelotoxicity will often be reached before therapeutic doses can be delivered to the tumor sites. Therefore, a preclinical model seems to be warranted, which would allow a more systematic study of the therapeutic efficacy and toxicity of systemic

treatment approaches in MTC. The TI' cell line used in this study, andit appears to be suitable model for this purpose: tumor take and growthrates were well reproduced; the cell line is heterozygous with respectto a typical mutation of the RET proto-oncogene; and it expresses thecharacteristic antigens of human MTCs, such as CEA and calcitonin,

as well as a range of other (neuro)peptides (15, 44). However, futurework will have to address, in more MTC cell lines, what influencedifferences in CEA expression and chemo- and radiosensitivity mayhave on the therapeutic results.

The anti-CEA MAb used for this study (clone F023C5) showedsome peculiarities as compared to other murine anti-CEA IgGs. First,its clearance from circulation was more rapid than that usually observed with murine IgGl . Overall half-lives were in a range more

typical for bivalent fragments (17, 24). Although a fairly rapid bloodclearance was observed in non-tumor-bearing mice as well, a goodcorrelation of the blood clearance to the tumor mass was found inxenografted animals. Although no formal studies were undertaken,which would correlate blood CEA levels or the amount of complexation between the injected antibody and circulating antigen to theclearance behavior of F023C5, hepatic clearance of immune complexes, formed between circulating antibody and antigen that are shedby the tumor, appears to be the most likely explanation, especially

because similar observations have been reported with anti-CEA andother antibodies that are directed against circulating antigens orepitopes (31—33).Accordingly, external scintigraphy showed an intense thyroid uptake (compare Fig. 1B) that could be prevented almost

completely by the addition of nonradioactive iodine to the animals'water supply. This clearly indicates metabolically liberated iodide andnot a potential cross-reactivity of the antibody with thyroid tissue asthe reason for the observed thyroid uptake.

The comparably modest tumor uptake values are most likely aconsequence of the low affinity (0.5 X 108 liters/mol) and rapid

clearance of the antibody. Future studies will show, in this context,whether antibodies of higher affinity may help to further enhance thetherapeutic efficacy.

Myelotoxicity is well known as a dose-limiting organ toxicity in

RIT (17, 27), and bone marrow support allowed for a dose intensification (17, 27), which is accordance with the clinical situation as well.

In accordance with earlier observations with ‘3'I-labeledIgG (24, 27),approximately 15 Gy to the blood (which is regarded as representative

5316

A0E‘a00>E15—

C150E

BI-0Eoe>E15—

C150E

100.0 1311-antl-CEA

10.0

1.0

500 1000 1500

1311FO23C5(pCi)

chemotherapy—0-- @j@tj 28

-@0- day 42

day 56

100.0

10.0

I .0

on April 5, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

IMPROVED TREATMENT OF MTC BY RICT

for the red marrow dose; Ref. 45) were identified as the MTD withoutartificial support, and approximately 28 Gy were identified as theMTh with bone marrow support. Interestingly, alkylating chemotherapeutic drugs (e.g., cisplatinum, dacarbazine, and streptozotocine) hadcomparable myelosuppressive side effects, whereas the myelosuppression caused by doxorubicin was very mild. Similar observations

have been reported for 5-FU in colorectal xenograft models by Blumenthal et a!. (46, 47). Severe 0! side effects (diarrhea, associatedwith a weight loss of over 20% in most animals) seemed to be doselimiting with doxorubicin, as has been reported previously in nudemice (48). This is also in accordance with clinical observations (1).However, the reported major long-term toxicity of doxorubicin inclinical settings is cardiac toxicity (e.g., congestive cardiomyopathy).Although some transient changes of CK levels in blood have beenobserved in this model, they do not seem to be sufficient to suggest theheart as the dose-limiting organ. However, chronic cardiac side effects, as are typical for doxorubicin in the clinical situation, may be

difficult to assess in a mouse model and may have escaped observation. Future work will have to address this problem in more detail.

Interestingly, doxorubicin and RIT with ‘311-labeledanti-CEA antibodies seemed to be comparably effective in this mouse model,which is in accordance with preliminary clinical data ( 12, 13, 36). Allother chemotherapeutic regimens tested were clearly inferior or evenonly borderline effective. Because the toxicity of both agents, doxorubicin and RIT, seemed to be complementary, we examined whether

combination approaches of these two therapeutic approaches may befeasible and whether they may enhance the therapeutic efficacy.Interestingly, very similar observations were reported earlier for combinations of 5-FUlleucovorin and RH' in a preclinical colorectalcancer model by Blumenthal et a!. (46).

Plotting tumor volume multiplication times against tumor doses(compare Fig. 7A) yielded a diagram that resembles a sort of invertedcell survival curve that is known from external beam radiation of cellsin culture (49). The resulting curves resemble the typical shouldercurves when cell survival is plotted against radiation doses in cellculture. This is not surprising because the tumor volume is a functionof the total number of cells and because the change in tumor volumeis a function of the number of viable cells that are able to proliferate.

Below tumor doses of 30—40Gy, the effects on tumor growth delaywere minimal. This shoulder appears rather high, even when compared to other solid tumors, such as colorectal cancer cell lines (whichare, per Se, relatively radioresistant as compared to other, e.g. , hematological, neoplasms). This behavior may indicate the human MTCcell line, TT, as fairly radioresistant, which, again, is in accordancewith the clinical situation ( 11).

Radioresistance of solid tumors can be attributed to a variety offactors, such as heterogeneous blood supply resulting in hypoxia, thecapacity to repair radiation-induced DNA damage, the repopulation ofcells after radiation, the predominance of a radioresistant phase of the

cell cycle, and the increased production of radical scavengers or theamplification of DNA repair genes (reviewed in Refs. 29, 49, and 50).Most studies of radiosensitizers were performed in conjunction withexternal beam radiation (50—52),but there have been several studieswith RIT as well (29, 47, 53—55).The use of radiosensitizers is anattractive concept, especially in Rif, in which tumor doses generally

are below 25 Gy (14, 39) in an attempt to reduce the radioresistanceof these tumors to improve the therapeutic efficacy. Radiosensitizerscan be divided into agents without antitumor effect alone and cytotoxic agents that do have an antitumor and radiosensitizing effect.Among the latter are several agents tried in this study as well, such ascisplatinum, 5-Ri, dacarbazine, and doxorubicin (37, 38, 49, 50).Because none of the other agents showed convincing antitumor effi

cacy by themselves, we focused our interest on the radiosensitizing

properties of doxorubicin.Like other anthracyclines, doxorubicin is a potent topoisomerase II

inhibitor, forming covalent topoisomerase-DNA complexes and preventing the enzyme from completing the religation portion of theligation-religation reaction (56). It is also intercalating between bp in

the DNA, causing single- and double-strand breaks (56). Furthermore,doxorubicin is, along with actinomycin D, known as a radiosensitizingagent (56—59). Clinical observations on the potentiation of the effects

of ionizing radiation in patients who had been treated with Adriamycm before or after external beam radiation (including “recallphenomena,― i.e. , reactivations of latent radiation effects in previously irra

diated fields by Adriamycin) had prompted several systematic studieson the mechanisms behind these effects (57—59). These studiesshowed that the doxorubicin effect primarily affects the shoulder, onlyslightly affecting the slope of cell survival curves ( 1, 57—59). This is

in perfect accordance with Fig. 7A, which suggests a “shift―of the

shoulder to lower dose levels, with the resulting dose-effect curveresembling the curve obtained with a more “radiosensitive―(colorectal) tumor. The analysis of these data according to the Loewe additivity model (28), as presented in Fig. 8 and Table 2, provides

evidence for a synergistic and not merely additive antitumor efficacy,which, again, clearly supports the view of doxorubicin as a radiosen

sitizing agent when it is used in conjunction with RIT. Although the

dosing schedule used in this study is based on dosimetric considerations (compare Fig. 7B), with the aim of exposing the tumor to theradiosensitizing agent at a time when the dose rate to other organs,especially to the radiosensitive red marrow, has fallen considerably,

more thorough and systematic studies are necessary to determine theoptimal dose and time scheduling between RIT and chemotherapy. Itis still interesting to note, however, that in the administration schemechosen, despite pronounced effects on the radiosensitivity of thetumor, effects on the red marrow were rather mild, although doxorubicin is known as potent radiosensitizer of hematopoietic stem cells aswell (60).

A combination strategy of RIT with doxorubicin, used to treathuman lung squamous cell carcinoma xenografts in nude mice, hasbeen reported previously by Desrues et a!. (48). They reported basicobservations that were similar to those presented in this study, such asa primarily GI toxicity of doxorubicin monotherapy associated withsevere weight loss, dose-limiting myelotoxicity of Rif, and enhancedtherapeutic efficacy of RICT combinations. However, no detailedanalysis on the underlying mechanisms, such as radioresistance of the

tumor cell line versus radiosensitizing properties of doxorubicin,additive versus synergistic mode of action, and so on, had beenperformed. In contrast, Desrues et a!. (48) interpreted their improvedtherapeutic results on the basis of the assumption that pre-RIT chemotherapy would lead to enhanced cell necrosis, with the consequenceof a better accessibility of the mainly intracytoplasmic antigen that isrecognized by the radioantibody; thus, higher tumor uptake values andradiation doses would result. Although no formal biodistributionstudies under doxorubicin challenge were performed here, major

changes in antibody biodistribution and dosimetry seem highly Un

likely because doxorubicin was not administered before 48 h after theradioantibody, which is a time point when the majority of the activityhas already cleared from the blood and the tumor uptake has reached

its apogee (compare Figs. lA and 7B). This timing seems to beparticularly important because doxorubicin has been shown to potentially decrease the tumor blood flow (61); this might compromise the

tumor uptake and, thus, the radiation dose, if the chemotherapeuticagent was administered before the radioantibody.

In summary, the human MTC model used for these studies seemsto be a suitable model for the preclinical testing of systemic thera

5317

on April 5, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

IMPROVED TREATMENT OF MTC BY RICF

peutic approaches in MTC. It seems to reflect the clinical situation(with respect to radiosensitivity and chemosensitivity, toxic side effects, and so on) fairly well. The relatively low chemosensitivity ofthe U cell line may be partially due to MDRI gene expression in thiscell line (62—66),which is a problem that is frequently encountered inthe clinical situation as well (1, 2, 36). Furthermore, the therapeuticresults of combined RIT with a radiosensitizing doxorubucin chemotherapy are encouraging. Further studies to confirm the basic observations of the present pivotal study, involving several MTC cell lineswith quantitatively different levels of CEA and/or MDRI gene expression, seem to be warranted. Furthermore, more systematic studiesare necessary to optimize the dosing (e.g. , combination schemes withvarying doses of Adriamycin) and scheduling of both therapeuticagents, to evaluate whether toxicity can be further minimized, whetherretreatment strategies may be feasible, and whether combined RICTmay improve the therapeutic efficacy in other (radioresistant) tumortypes as well. Clinical studies on the combination of RIT and chemotherapy in metastatic MTC are ongoing.

ACKNOWLEDGMENTS

We thank Dr. D. M. Goldenberg (Garden State Cancer Center, Belleville,NJ) for providing us with the GW-39 human colon cancer cell line. The expertadvice of Prof. Dr. M. Oellerich and Dr. E. Wieland (Department of ClinicalChemistry, Georg-August-University, Gottingen, Germany) on blood count

determinations and blood chemistry is gratefully acknowledged. Furthermore,

we thank Drs. K. Nebendahl and 0. Schunck for their valuable advice onanimal care. Last but not least, we are indebted to Dr. W. Hoppner (Hamburg,Germany) for performing the genetic analysis on the TI' cell line, with respect

to mutations in the RET proto-oncogene.

REFERENCES

1. Dc Vita, V. T., Hellman, S., and Rosenberg, S. A. (eds.). Cancer: Principles andPractice of Oncology, Ed. 5. Philadelphia: Lippincott-Raven, 1997.

2. Raue,F. (ed). MedullaryThyroidCarcinoma.Berlin:Springer-Verlag,1992.3. Pacini, F., Fontanelli, M., Fugazzola. L., Elisei, R., Romei, C., Di Coscio, G., Miccoli,

P.. and Pinchera, A. Routine measurement of serum calcitonin in nodular thyroiddiseases allows the preoperative diagnosis of unsuspected sporadic medullary thyroidcancer. J. Clin. Endocrinol. Metab., 78: 826—829, 1994.

4. Bergholm, U., Bergstrom, R., and Ekbom, A. Long-term follow-up of patients withmedullary carcinoma of the thyroid. Cancer (Phila.), 79: 132—138,1997.

5. Tisell, L. E., Hansson, G., Jansson, S., and Salander, H. Reoperation in the treatmentof asymptomatic metastasizing medullary thyroid carcinoma. Surgery (St. Louis), 99:60—66,1986.

6. Tisell, L. E., Dilley. W. G., and Wells, S. A. Progression of postoperative residualmedullary thyroid carcinoma as monitored by plasma calcitonin levels. Surgery (St.Louis), 119: 34—39,1996.

7. Samaan, N. A., Schultz, P. N., and Hickey, R. C. Medullary thyroid carcinoma:prognosis of familial versus sporadic disease and the role of radiotherapy. J. Clin.Endocrinol. Metab., 67: 801—808,1988.

8. van Heerden, J. A., Grand, C. S., Gharib, H., Hay, I. D., and llstrup D. M. Long-termcourse of patients with persistent hypercalcitoninemia after apparent curative primarysurgery for medullary thyroid cancer. Ann. Surg., 212: 395—401, 1990.

9. Behr, T. M., Gratz, S., Markus, P. M., Dunn, R. M., Htifner, M., Becker, H., andBecker, W. Enhanced bilateral somatostatin receptor expression in mediastinal lymphnodes (“chimneysign―)in occult metastatic medullary thyroid cancer: a typical signof tumour manifestation? Eur. J. NucI. Med., 24: 184—191,1997.

10. Behr, 1. M., Gratz. S., Markus, P. M., Dunn, R. M., Htifner, M., Schauer, A., Fischer,M., Munz, D. L., Becker, H., and Becker, W. S. Anti-CEA antibodies s'ersussomatostatin-analogs in detection of metastatic medullary thyroid cancer: are CEAand somatostatin-receptor expression prognostic factors? Cancer (Phila.), in press,1997.

I I. Simpson, W. J., Sutcliffe, S. B., and Gospodarowicz, M. K. The thyroid. In: Cox,J. D. (ed). Moss' Radiation Oncology: Rationale, Technique, Results, Ed. 7, pp.280—304.St. Louis:Mosby,1994.

12. Juweid, M. E., Sharkey, R. M., Behr, T., Swayne, L. C., Rubin, A. D., Hanley, D.,Markowitz, A., Siegel, J., and Goldenberg, D. M. Targeting and initial radioimmunotherapy of medullary thyroid cancer with ‘3'I-labeled monoclonal antibodies tocarcinoembryonic antigen. Cancer Res., 55 (Suppl.): 5946s—595ls, 1995.

13. Juweid, M. E., Sharkey, R. M., Behr, 1., Swayne, L. C., Herskovic, T., Pereira, M.,Rubin, A. D., Hanley, D., Dunn, R. M., Siegel, J., and Goldenberg. D. M. Radioimmunotherapy of medullary thyroid carcinoma with 311-labeled anti-CEA antibodies.J. NucI.Med.,37: 875—881,1996.

14. Behr, T. M., Sharkey. R. M., Juweid, M. E., Dunn, R. M., Siegel, J. A., and

Goldenberg, D. M. Variables influencing tumor dosimetry in radioimmunotherapy ofCEA-expressing cancers with anti-CEA and anti-mucin monoclonal antibodies.J. NucI. Med., 38: 409—418,1997.

15. American Type Culture Collection. ATCC Cell Lines and Hybridomas, Ed. 8, p. 174.Rockville, MD: American Type Culture Collection, 1994.

16. Goldenberg, D. M., Witte, S., and Elster, K. A new human tumor serially transplantable in the golden hamster. Transplantation (Baltimore), 4: 760—764,1966.

17. Behr, T. M., Sharkey, R. M., Sgouros, G., Blumenthal, R. D., Dunn, R. M., Kolbert,K., Griffiths, G., Siegel, 3. A., Becker, W. S., and Goldenberg, D. M. Overcomingnephrotoxicity of radiometal-labeled immunoconjugates: improved cancer therapy ina nude mouse model in relation to the internal radiation dosimetry. Cancer (Phila.), inpress, 1997.

18. Behr, T., Becker, W., Bair, H. J., Klein, M., Stühler,C. M., Cidlinsky, K. P., Scheele,J. R., and Wolf, F. G. Comparison of complete versus fragmented @“Tc-labeledanti-CEA monoclonal antibodies for immunoscintigraphy in colorectal cancer.J. NucI. Med., 36: 430—441, 1995.

19. Behr, T., Becker, W., Hannappel, E., Goldenberg, D. M., and Wolf F. Targeting ofliver metastases ofcolorectal cancer with IgG, F(ab')2, and Fab' anti-CEA antibodieslabeled with @‘“Tc:the role of metabolism and kinetics. Cancer Res., 55 (Suppl.):5777s—5785s, 1995.

20. Primus, F. J., Newell, K. D., Blue, A., and Goldenberg, D. M. Immunological heterogeneity of carcinoembryonicantigen: antigenicdeterminantson carcinoembryonicantigendistinguished by monoclonal antibodies. Cancer Res., 43: 686-692, 1983.

21. Weadock, K. S., Sharkey, R. M., Varga, D. C., and Goldenberg, D. M. Evaluation ofa remote radioiodination system for radioimmunotherapy. J. NucI. Med., 31: 508—511, 1990.

22. Behr,T. M., Sharkey,R. M., Juweid,M. E., Dunn, R. M., Siegel,J. A., andGoldenberg, D. M. Factors influencing the pharmacokinetics, dosimetry, and diagnostic accuracy of radioimmunodetection and radioimmunotherapy of CEA-expressing tumors. Cancer Rca., 56: 1805—1816,1996.

23. Behr, 1. M., Sharkey, R. M., Juweid, M. E., Blumenthal, R. D., Dunn, R. M., Bair,H. J., Griffiths, G. L, Wolf, F. G., Becker, W. S., and Goldenberg D. M. Reductionof the renal uptake of radiolabeled monoclonal antibody fragments by cationic aminoacids and their derivatives. Cancer Rca., 55: 3825—3834, 1995.

24. Sharkey, R. M., Motta-Hennessy, C., Pawlyk, D., Siegel, J. A., and Goldenberg,D. M. Biodistribution and radiation dose estimates for yttrium- and iodine-labeledmonoclonal antibody IgG and fragments in nude mice bearing human colonic tumorxenografts. Cancer Res., 50: 2330—2336, 1990.

25. Siegel, J. A., and Stabin, M. G. Absorbed fractions for electrons and f3 particles in

small spheres. J. NucI. Med., 29: 803, 1988.26. Hui, 1. E., and Poston, J. W. A model of the circulating blood for use in radiation dose

calculations. In: Proceedings of the International Conference on Radiation Dosimetryand Safety, pp. 151—168.Taiwan: 1987.

27. Blumenthal, R. D., Sharkey, R. M., Forman, D., Wong, G., Hess, J., and Goldenberg,D. M. Improved experimental cancer therapy by radioantibody dose intensification asa result of syngeneic bone marrow transplantation. Exp. Hematol., 23: 1088—1097,1995.

28. Greco, W., Bravo, G., and Parsons, J. The search for synergy: a critical review from

a response surface perspective. Pharmacol. Rev., 47: 331—385,1995.29. Kievit, E., Pinedo, H. M., SchlQper, H. M. M., and Boven, E. Addition of cisplatin

improves the efficacy of ‘3'I-labeledmonoclonal antibody 323/A3 in experimentalhuman ovarian cancer. Int. J. Rad. Oncol. Biol. Phys., 38: 419—428,1997.

30. Berenbaum, M. What is synergy? Pharmacol. Rev., 41: 93—141,1989.31. Pimm, M. V. Circulating antigen: bad or good for immuno-scintigraphy? NucI. Med.

Biol.,22:137—145,1995.32. Davies, Q., Perkins, A. C., Frier, M., Watson, S., Lalani, E-N., and Symonds, E. M.

The effect of circulating antigen on the biodistribution of the engineered humanantibody hCFMO1 in a nude mice model. Eur. J. NucI. Med., 24: 206—209,1997.

33. Beatty, B. G., Beatty, J. D., Williams, L. E., Paxton, R. 3., Shively, J. E., andO'Connor-Tressel, M. Effect of specific antibody pretreatment on liver uptake ofI I ‘In-labeled anticarcinoembryonic antigen monoclonal antibody in nude mice bear

ing colon cancer xenografts. Cancer Res., 49: 1587—1594,1989.34. Benker, G., and Reinwein, D. Ergebnisse der chemotherapie des schilddrtisenkarzi

noms. Dtsch. Med. Wochenschr., 108: 403—406, 1983.35. Schlumberger, M., Abdelmoumene, N., Delisle, M. J., Couette, J. E., and Groupe

d'Etude des Tumeurs a Calcitonine (GETC). Treatment of advanced medullarythyroid cancer with an altemating combination of 5-FU-streptozotocin and 5-FUdacarbazine. Br. J. Cancer, 71: 363—365,1995.

36. Gottlieb, J. A., and Hill, C. S. Chemotherapy of thyroid cancer with Adriamycin.N. EngI. 1. Med., 290: 193—197,1974.

37. Cassady, J. R., Richter, M. P., Piro, A. J., and Jaffe, N. Radiation-Adriamycin

interactions: preliminary clinical observations. Cancer (Phila.), 36: 946—949,1975.38. Belli, J. A., and Piro, A. J. The interaction between radiation and Adriamycin damage

in mammalian cells. Cancer Res., 37: 1624—1630, 1977.39. Behr, T. M., Goldenberg, D. M., and Becker, W. S. Radioimmunotherapy of solid

tumors: a review “ofmice and men.―Hybdridoma, 16: 101—107,1997.40. Guilloteau, D., Baulieu, J. L., and Besnard, J. C. Medullary thyroid carcinoma

imaging in an animal model: use of radiolabeled anticalcitonin F(ab')2 and metsiodobenzylguanidine. Eur. J. NucI. Med., 11: 198—200,1985.

41. Mendelsohn, G.. Wells, S. A., Jr., and Baylin, S. B. Relationship of tissue carcinoembryonic antigen and calcitonin to tumor virulence in medullary thyroid carcinoma.Cancer (Phila.), 54: 657—662,1984.

42. Busnardo, B., Girelli, M. E., Simioni, N., Nacamulli, D., and Bosetto, E. Nonparallelpatterns of calcitonin and carcinoembryonic antigen levels in the follow-up ofmedullary thyroid carcinoma. Cancer (Phila.). 53: 278—285, 1984.

43. Bardiès, M., Bardet, S., Faivre-Chauvet, A., Peltier, P., Douillard, J-Y., Mahé,M.,

5318

on April 5, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

IMPROVED TREATMENT OF MTC BY RICT

Fiche, M., Lisbona, A., Giacalone, F., Meyer, P., Gautherot, E., Rouvier, E., Barbet,J,, and Chatal, J. F. Bispecific antibody and iodine-l31-labeled bivalent haptendosimetryinpatientswithmedullarythyroidor small-celllungcancer.J. NucI.Med.,37: 1853—1859,1996.

44. Zabel, M. and Grzeszkoviek, F. Characterisation of thyroid medullary carcinoma ‘ITcell line.Histol.Histopathol.,12: 283—289,1997.

45. Sgouros, G. Bone marrow dosimetry for radioimmunotherapy: theoretical considerations. J. NucI. Med., 34: 689—694,1993.

46. Blumenthal, R. D., Sharkey, R. M., Natale, A. M., Kashi, R., Wong, G., andGoldenberg, D. M. Comparison of equitoxic radioimmunotherapy and chemotherapyin the treatment of human colonic cancer xenografts. Cancer Res., 54: 142—151, 1994.

47. Blumenthal, R. D., Sharkey, R. M., Gold, D. V., Osorio, L., and Goldenberg, D. M.Multimodal experimental therapy of a s.c. colonic cancer xenograft using radioimmunotherapy (RAff) and chemotherapy. 3. NucI. Med., 37: l69P, 1996.

48. Desrues, B., Brichory, F., Lena, H., Bourguet, P., Delaval, P., Toujas, L, and DazordL. Treatment of human lung cancer xenografts with a combination of ‘3'l-labelledmonoclonal antibody Po66 and doxorubicin. Cancer Immunol. Immunother., 43:269—274, 1996.

49. Hall, E. J. Radiobiology for the Radiologist. Philadelphia: J. B. Lippincott Co., 1994.50. Walden, T. L., Jr., and Farzaneh, N. K. Biochemistry of Ionizing Radiation. New

York: Raven Press, 1990.51. Coleman, C. N., and Turrisi, A. 1. Radiation and chemotherapy sensitizers and

protectors. Crit. Rev. Oncol. Hematol., 10: 225—252,1990.52. Kirichenko, A. V., Rich, T. A., Newman, R. A., and Travis, E. L. Potentiation of

murine MCa-4 carcinoma radioresponse by 9-amino-20(S)-camptothecin. CancerRes., 57: 1929—1933,1997.

53. Chalandon, Y., Mach, J. P., Pélégrin,A., Folli, S., and Buchegger, F. Combinedradioimmunotherapy and chemotherapy of human colon carcinoma xenografted nudemice, advantages and limitations. Anticancer Res., 12: 1131—1140, 1992.

54. Remmenga. S., Colcher, D., Gansow, 0., Pippen, G., and Raubitschek, A. Continuousinfusion chemotherapy as a radiation-enhancing agent for yttrium-90-radiolabeledmonoclonal antibody therapy of a human tumor xenograft. Gynecol. Oncol., 55:115—122,1994.

55. Roffler, S. R., Chan, J., and Yeh, M. Y. Potentiation of radioimmunotherapy byinhibition of topoisomerase I. Cancer Res., 54: 1276—1285, 1994.

56. Corbett, A. H., and Osheroff, N. When good enzymes go bad: conversion oftopoisomerase II to a cellular toxin by antineoplastic drugs. Chem. Res. Toxicol., 6:585—597,1993.

57. Donaldson, S. S., Glick, J. M., and Wilbur, J. R. Adriamycin activating a recall

phenomenon after radiation therapy. Ann. Intern. Med., 81: 407—408,1974.58. Bases, R. E. Modification of the radiation response determined by single-cell tech

nics: actinomycin D. Cancer Res., 19: 1223—1229,1959.59. Piro, A. J., Taylor, C. C., and Belli, J. A. Interaction between radiation and drug

damage in mammalian cells: I. Delayed expression of actinomycin D/x-ray effects inexponential and plateau phase cells. Radiat. Res., 63: 346—362, 1975.

60. Hellman, S., and Hannon, E. Effects of Adriamycin on the radiation response ofmurine hematopoietic stem cells. Radiat. Res., 67: 162—167,1976.

61. Durand, R. E., and LePard, N. E. Tumour blood flow influences combined radiationand doxorubicin treatments. Radiother. Oncol., 42: 171—179,1997.

62. Yang, K-P., and Samaan, N. A. Lethal efficacy of doxorubicin on human medullarythyroid carcinoma cells in vitro. Anticancer Res., 8: 245—248,1988.

63. Larsson, R., and Nygren, P. Verapamil and cyclosporin A potentiate the effects ofchemotherapeutic drugs in human medullary thyroid carcinoma ‘ITcell line notexpressing the 170 kDa P-glycoprotein. Cancer Lett.. 54: 125—131, 1990.

64. Yang, K-P., Liang, Y-F., and Samaan, N. A. Intrinsic drug resistance in a humanmedullary thyroid carcinoma cell line: association with overexpression of mdrl geneand low proliferation fraction. Anticancer Res., I 1: 1065—1068,1991.

65. Massart, C., Gibassier, J., Raoul, M., Pourquier, P.. Leclech, G., Robert, J., andLucas, C. Cyclosporin A, verapamil. and 59788 reverse doxorubicin resistance ina human medullary thyroid carcinoma cell line. Anti-Cancer Drugs, 6: 135—146,1995.

66. Massart, C., Gibassier, J., Lucas, C., Pourquier. P.. and Robert, J. Expression du geneMDR1 dans cinq lignéeshumaines de cancer mCdullairede Ia thyroIde et reversionde Ia résistancea Ia doxorubicine par Ia ciclosporine A et Ic vérapamil.Bull. Cancer,83: 39—45,1996.

5319

on April 5, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

1997;57:5309-5319. Cancer Res Thomas M. Behr, Erik Wulst, Sven Radetzky, et al. Radiolabeled Antibodies in a Radioresistant Tumor TypeDoxorubicin Potentiates the Therapeutic Efficacy ofMouse Model by Combined Radioimmunochemotherapy: Improved Treatment of Medullary Thyroid Cancer in a Nude

Updated version

http://cancerres.aacrjournals.org/content/57/23/5309

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/57/23/5309To request permission to re-use all or part of this article, use this link

on April 5, 2021. © 1997 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/content/57/23/5309http://cancerres.aacrjournals.org/cgi/alertsmailto:pubs@aacr.orghttp://cancerres.aacrjournals.org/content/57/23/5309http://cancerres.aacrjournals.org/