(CANCER RESEARCH 49, 2639-2644, May 15, 1989]

Nature of the Bifunctional Chelating Agent Used for Radioimmunotherapy withYttrium-90 Monoclonal Antibodies: Critical Factors in Determining in Vivo

Survival and Organ ToxicityRobert W. Kozak,1 Andrew Raubitschek, Saed Mirzadeh, Martin W. Brechbiel, Richard Junghaus, Otto A. Gansow,

and Thomas A. WaldmannDivision ofCytokine Biology, Center for Biologies Evaluation and Research, FDA [R. W. K.J, Radiation Oncology Branch [A. R., S. M., M. W. B., O. A. G.J, andMetabolism Branch [R. ]., T. A. WJ, National Cancer Institute, NIH, Bethesda, Maryland 20892

ABSTRACT

One factor that is critical to the potential effectiveness of radioimmu-notherapy is the design of radiometal-chelated antibodies that will be

stable in vivo. Stability in vivo depends on the condition that both theelidale linkage and radiolabeling procedures not alter antibody specificityand biodistribution. In addition, synthesis and selection of the chelatingagent is critical for each radiometal in order to prevent inappropriaterelease of the radiometal in vivo. In the present study, we compare thein vivo stability of seven radioimmunoconjugates that use differentpolyaminocarboxylate chelating agents to complex yttrium-88 to themouse anti-human interleukin-2 receptor monoclonal antibody, anti-Tac.Chetate linkage and radiolabeling procedures did not alter the immuno-specificity of anti-Tac. In order to assess whether yttrium was inappropriately released from the chelate-coupled antibody in vivo, iodine-131-labeled and yttrium-88 chelate-coupled antibodies were simultaneously

administered to the same animals to correlate the decline in yttrium andradioiodinated antibody activity. The four stable yttrium-88 chelate-coupled antibodies studied displayed similar iodine-131 and yttrium-88activity, indicating minimal elution of yttrium-88 from the complex. Incontrast, the unstable yttrium-88 chelate-coupled antibodies had serumyttrium-88 activities that declined much more rapidly than their iodine-131 activities, suggesting loss of the radiolabel yttrium-88 from thechelate. Furthermore, high rates of yttrium-88 elution correlated withdeposition in bone. Four chelating agents emerged as promising immu-

notherapeutic reagents: isothiocyanate benzyl DTPA and its derivatives1B3M, MX, and 1M3B. All four isothiocyanate agents showed prolongedyttrium-88 vascular survival which was essentially identical to that oftheir iodine-131 activity with only minimum accumulation (1.4-1.8%/g)of the yttrium-88 injected dose into bone. Thus, these four chelatingagents were very stable in vivo and suitable for yttrium-monoclonal

antibody radioimmunotherapy.

INTRODUCTION

MAb2 directed against tumor-associated antigens have been

used as immunoimaging and itnmunotherapeutic agents in human cancer. Unfortunately, there have been only 23 partial andthree complete remissions reported in the 185 patients includedin the 25 clinical trials involving 13 unconjugated murine MAb(1). There have been a number of explanations for the lowtherapeutic efficacy observed. A major one is that many of theMAb employed are not cytocidal or cytostatic against neoplasticcells in humans. The limited and sometimes transient success

Received 10/4/88; revised 2/15/89; accepted 2/16/89.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 inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

' To whom requests for reprints should be addressed at Building 29A, Room2A11, HFB-500, 8800 Rockville Pike, Bethesda, MD 20892.

2The abbreviations used are: MAb, monoclonal antibody; IL-2, interleukin 2;

DTPA, diethylenetriaminepentaacetic acid; CA, chelating agent; BSA, bovineserum albumin; FCR, fractional catabolic rate; FER, fractional elution rate;SCN-BZ-DTPA, isothiocyanate benzyl diethylenetriaminepentaacetic acid;CaDTPA, cyclic dianhydride of DTPA; 1B-EDTA, l-(p-isothicyanatoben-zyl)ethylenediaminetetraacetic acid; 2B-DTPA, 2-(p-isothiocyanatobenzyl)-DTPA; 1M3B-DTPA, 2-(p-isothiocyanatobenzyl)-4-methyl-DTPA; MX-DTPA,l-(2)-methyl-4-SCN-BZ-DTPA; 1B3M-DTPA, H/»-isothiocyanatobenzyl)-3-methyl-DTPA.

of unmodified MAb in eliminating the target tumor cells hasprovided an impetus for developing radionuclide-chelated antibodies as immunotherapeutic agents for use in the treatmentof lymphomas, leukemia, graft rejection, and autoimmune disease. The results obtained with the available radiometals che-

lated to MAb have been encouraging; however, it is clear thatfurther development is required before radiolabeled antitumorantibodies can achieve their full potential as radioimmuno-

therapeutic agents.There are a number of components that must be considered

in designing optimal radioimmunotherapeutic reagents for invivo studies (2-5). Among the issues to be addressed is theselection of the MAb that serves to carry the radionuclide tothe tumor target. The MAb is selected based on the distributionof its antigenic target and on the specificity and binding affinityof the antibody to its target. In the present study, we used theanti-Tac MAb, which reacts with the M, 55,000 peptide ofprimate but not murine IL-2 receptors. This receptor is notexpressed on resting cells but is expressed on the surface of anumber of different forms of activated or malignant B-cells,monocytes, and T-cells, including those from the human T-celllymphotrophic virus type I-associated adult T-cell leukemia (6).

A second component to consider is the nature of the radionuclide used. The radionuclide is selected based on its physical,chemical, and biological properties. An optimal radionuclideshould be routinely available, easy to couple to the MAb, shouldhave a short range, high energy, and abundant particle emission,a stable daughter product, and an appropriate physical half-lifeto selectively eliminate the target neoplastic tissue while sparingnormal tissues.

A third and pivotal issue in designing an optimal radioimmunotherapeutic reagent is the choice of the CA used to couplethe radionuclide to the MAb. An ideal CA fulfills the followingcriteria: (a) its addition should not alter the specificity or thebinding affinity of the MAb to its antigenic target; (b) itsaddition to the MAb should not otherwise damage the antibodyand thus alter its rates of catabolism or patterns of tissuedistribution; (<•)it should hold the radiometal tightly so that

there is no premature elution of the radioisotope from theMAb-CA complex in vivo; and (d) it should help clear theradiometal following catabolism of the MAb-CA-radiometalcomplex. These criteria are required in order to maintain theadvantage provided by the specificity of the MAb utilized. Themost common cause of failure of the CA is that it does not linkand securely hold the radionuclide to the antibody. As a consequence, there is considerable dissociation of the radionuclidein vivo from the MAb-CA complex prior to delivery of theseagents to the surface of the tumor cells. The free radionuclidesdo not contribute to the production of the desired radiothera-peutic effects but, rather, are delivered to normal tissues wherethey cause toxicity. Therefore, in order to develop, compare,and select CAs that will be of greatest value in radioimmuno-

2639

on May 22, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CHELATING AGENTS FOR YTTRIUM RADIO1MMUNOTHERAPY

therapy, one should define the metabolism of the MAb itself,of the M Ab-CA and M Ab-CA-radiometal complex.

In the present study, we have compared a series of backbone-substituted isothiocyanate benzyl diethylenetriaminepentaa-cetic acid (SCN-BZ-DTPA) bifunctional CAs to more conventional CAs in terms of their ability to bind yttrium-88 to thehuman IL-2 receptor MAb, anti-Tac. The addition of alkylgroups to the backbone of DTPA conferred increased stabilityto the subsequent metal complex (7). Yttrium-88 was utilizedto model the metabolism of this element, since we anticipateusing the ^-emitting yttrium-90 chelates of anti-Tac in thetherapy of patients with Tac-expressing neoplastic diseases.The different CAs analyzed were evaluated in terms of thecriteria for an ideal CA outlined earlier. The effect of an addedCA on the capacity of the MAb to bind to its antigenic targetwas examined by comparing the ability of radioiodinated amiTac-CA to bind to the cell lines with the ability of radioiodinated unmodified anti-Tac to combine to the same targets. Theeffects of the chemical procedures used to couple the CA withthe antibody on the rates of catabolism and patterns of distribution of the antibody were assessed by comparing the metabolism of iodine-131-labeled anti-Tac-CA with the metabolismof unmodified u'I-labeled anti-Tac in blood metabolic clearance

studies. The capacity of each bifunctional CA to hold securelythe radionuclide to the MAb was defined by comparing theplasma survival of yttrium-88 chelated by the MAb-CA withI31l-labeled MAb-CA. With such studies, the elution rates ofyttrium-88 from the different CAs could be defined. Finally,the quantity of yttrium-88 in various organs was compared tothat of iodine-131 to define the distribution of yttrium releasedfrom the CA in vivo. SCN-BZ-DPTA and its derivatives 1M3B,MX, and 1B3M emerged as very stable CAs for use with yttriumand anti-Tac in vivo and thus are good candidates for furthercharacterization as radioimmunotherapeutic CAs with yttrium.

MATERIALS AND METHODS

MAb Anti-Tac. The MAb anti-Tac is a murine Igd»antibody thatis specific for the human p55 IL-2 receptor chain (8). Anti-Tac waspurified by DEAE-52 cellulose (Whatman, Kent, England) followed bysodium sulfate precipitation. Purity was established by immunoelectro-phoresis and sodium dodecyl sulfate-polyacrylamide gel electrophoresiswith silver staining.

Coupling of CA to and Yttrium Labeling of Anti-Tac. CA synthesisand coupling of CAs to MAb have, in part, been previously described(9). CAs used are listed in Table 1 and were CaDTPA, 1B-EDTA,SCN-BZ-DTPA, 2B-DTPA, 1M3B-DTPA, MX-DTPA, and 1B3M-DTPA. For accurate measurement of the number of chelates perantibody, all chelating agents were I4C labeled (9, 10). The CA/MAb

ratios for the present study ranged from 0.53 to 1.60 (Table 2). Adetailed description of the lengthy synthesis of all the CAs is providedby Brechbiel (10). Yttrium-88 was purchased from Oak Ridge NationalLaboratories (Oak Ridge, TN).

Anti-Tac was labeled with yttrium-88 by combining -100 M'of anti-Tac-CA (l mg, pH 4.5) with 50 ¿il88Y-acetate (pH 4.0) to yield a finalsolution at pH 4.2. Following a 20-min reaction time at room temperature, 10 n\ of 2 M sodium acetate and 2 /¿Iof 10~2M N32EDTA were

added. After 5 min, the samples were purified on a TSK-4000 sizeexclusion column using MES/C1~ buffer at pH 6.2 [20 mM in MES,

150 mM in NaCl 2-(ALmorpholino)ethanesulfonic acid] (Calbiochem-Behring, La .lolla. CA). High-pressure liquid chromatography wasoperating at a pressure of ~20 bars, with a flow rate of 1.0 ml/min.Under the above conditions, proteins with the M, of 1.5 x 10s g were

eluted with a retention time of 11.5 min with full width at half maximum= 1.0 min. In all cases, better than 80% of 88Y-labeled anti-Tac was

recovered in 2 ml of buffer, with excellent separations from highermolecular weight aggregates and unreacted yttrium-88 (EDTA, citrate,

Table 1 Bifunctional chelating agents

Structure Name"

COiH COiH

Ca-DTPA*

1B-EDTA

2B-DTPA

CO.H CO.H

1B-DTPA(SCN-BZ-DTPA)

1M3B-DTPA

CO,H CO,H

MX-DTPA

1B3M-DTPA

°See "Materials and Methods" for full name of chelating agents.* CaDTPA must use one of its carboxyl groups to form an amide linkage with

the Mab.

and oxychlorocomplexes) having retention times of R, ~ 16 min.Iodine-131 Labeling of CA-coupled and Unmodified Anti-Tac. Anti-

Tac was radioiodinated using the iodine monochloride technique ofMcFarlane (11) as previously described (12). Briefly, l mg of CA-coupled or unmodified anti-Tac was added to 0.5 ml of borate buffer,pH 8.0 (0.1 M borate in 1.0 M NaCl). If necessary, this solution wasadjusted to pH 8.0. In another tube, 1 mCi of Na'3'I (specific activity,

approximately 10 Ci/mg, cat. no. NEZ-035A, New England Nuclear/Dupont, Boston, MA) was added to 0.5 ml of borate buffer. Approximately 0.15 ml of a 1:100 (diluted with l M NaCl) dilution of iodinemonochloride, ICI, stock (0.033 M ICI) was added to the radioiodide.The radioactive solution was immediately mixed and added to theprotein solution, mixed again, and rapidly placed on a G25 column(PD10; Pharmacia, Piscataway, NJ), prc-equilibrated with 30 ml ofphosphate buffered saline (pH 7.2). Fractions of 0.5 ml were collectedin tubes containing 0.5 ml of a 0.5% BSA solution, and the efficiencyof labeling was determined. A specific activity of about 1 MCi/10 MÕÕwas achieved. If necessary, the reaction was repeated with less radioactivity and more protein to approach this specific activity.

Immunoreactivity of Labeled Anti-Tac. The percentage of radiola-beled anti-Tac that retained immunoreactivity was assessed by incubation of 5-50 ng of labeled antibody with 2-20 x 10' HUT-102 cells, aTac-positive cell line. The binding assay was performed in RPMI 1640supplemented with 1% BSA, 25 mM 4-(2-hydroxyethylene)-l-pipera-zineethanesulfonic acid buffer, 0.01% sodium azide and 100 //«/mlofhuman IgG to reduce nonspecific binding. Binding to MOLT 4, a Tac-

2640

on May 22, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CHELATING AGENTS FOR YTTRIUM RADIOIMMUNOTHERAPY

Table 2 Radionuclide labeling ofCA-coupled anti-TacUnmodified and seven différentCA-coupled anti-Tac preparations were labeled

with Iodine-131 using the iodine monochloride method or Yttrium-88 (see"Materials and Methods"). Bindability was assessed on IL-2 receptor-positiveand -negative cell lines. The bindability of a 3H-labeled anti-Tac internal controlwas only 60% of maximum in the assays for CaDTPA and 2B-DTPA; therefore,the immunoreactivity of the CA-coupled antibodies was similar to this value.

CANoneCa-DTPA1B-EDTA2B-DTPACA/antibodyratio0.531.281.14RadionuclideIodine-131Iodine-131

Yttrium-88Iodine-131

Yttrium-88Iodine-131

Yttrium-88Specific

activityGiCi/<ig)0.340.07

0.030.06

0.100.39

0.016Bindability

(%)7768

5884

706867

SCN-BZ-DTPA 1.25 Yttrium-88 0.08 82

1M3B-DTPAMX-DTPA1B3M-DTPA1.601.201.41Iodine-131Yttrium-88Iodine-131

Yttrium-88Iodine-131

Yttrium-880.07

0.080.11

0.100.17

0.11SO

9182

908874

negative line, or in the presence of excess unlabeled anti-Tac was usedas a measure of nonspecific binding. Assays were performed at roomtemperature or at 0°Cfor 1-2 h. Cell-bound counts were then deter

mined by pelleting cells 12,000 x g for 3 min and counting by standardmethods. The extent of binding of a 3H-labeled anti-Tac preparationwas used in each assay as an internal control. Anti-Tac was 3H-labeled

to high specific activity by a sodium borohydride reductive methylationtechnique as described by Tack and his coworkers (13).

In Vivo Distribution of Radiolabeled Anti-Tac. Athymic nude mice(6-week-old female, six per group) were coinjected in their tail veinwith 1 /iCi/mouse of I3'l-labeled anti-Tac-CA and 1 nC\ of anti-Tac-CA-88Y premixed in a total volume of 0.1 ml. One group of animalsreceived only 1 nC'i of I3ll-labeled unmodified anti-Tac. Solutions were

filtered with a 0.22 urn GV (Millipore, Bedford, MA) low protein-binding filter. Blood samples were collected 3-5 min postinjection andat various times over the next 4 days by tail vein bleeding, followed byeuthanasia on Day 5. Organs were collected, rinsed, blotted dry, andweighed on an analytical balance. The ends of the femur heads wereclipped off, and the bone marrow was flushed and counted separately.Blood and organ samples were then counted on a gamma-counter[Nal(Tl), well type 5500; Packard Instruments, Downers Grove, IL]along with reference samples for each of the injected doses from all thelabeled antibodies. Data was processed by background subtraction,decay, and crossover correction. Under our counting conditions, iodine-131 had a 6% and yttrium-88 a 21% spillover into each other's counting

windows. Organ data was calculated as the percentage of the injecteddose per gram of tissue, and blood samples were normalized to theirinitial time point taken 3-5 min postinjection.

Metabolism of Radiolabeled Anti-Tac. Blood clearance data wereprocessed as previously described (12, 14). Intravascular activity wasexpressed as a fraction of the initial injected dose (5-min time point)and plotted as a function of time on a semilogarithmic scale. Thebiological half-life (t\n) of the labeled antibody was determined graphically from the time point equal to one-half the fraction of the injecteddose identified by extrapolating the blood clearance exponential curveto time zero. The fraction of the intravascular protein pool catabolized

per day (FCR) was determined according to Matthews (15) as:

FCR =_b\

_b„

This approach involves the graphic analysis of the plasma radiola-

beled protein clearance curve. The original plasma curve was plottedon semilogrithmic paper, and its straight linear terminal portion wasextrapolated to the ordinate to obtain the intercept C\. The slope ofthis line is -61. By subtracting the extrapolated line from the originalcurve, a new curve is obtained from which the slope and intercept valuesCi and -¿»2are obtained in the same manner as with the original curve.A third "peeling" yields C} and —¿�b}.Thus, the original curve may bedescribed by three exponentials, the slopes —¿�b¡,—¿�¿»2,and —¿�bj, with

corresponding intercepts Ct, C2, and C3. These values for slopes andintercepts are used to determine the metabolic parameters for theprotein.

The apparent FCR of yttrium-88 (FCR-88Y) was determined fromthe serum clearance curves for yttrium-88-cheIated antibodies using atechnique identical to that used for the simultaneous radioiodinatedantibodies. The FER of yttrium-88 from the antibody chelate wasdetermined from the difference between the apparent FCR of 88Y-labeled antibodies and the rates of MAb-CA that were radioiodinated(FCR-'3'I) as follows: FER equals FCR-88Y minus FCR-'31I.

RESULTS

Immunospecificity Following Conjugate Assembly. The initialstudies aimed to determine if either the addition of the CA orthe incorporation of the yttrium label altered the capacity oflabeled M Ah to bind to its antigenic target, the Tac peptide. Inthese in vitro studies of immunospecificity, we contrasted thebinding of 13ll-labeled unmodified anti-Tac, I3'l-labeled anti-Tac-CA, and anti-Tac-CA-88Y complexes to the Tac-expressingHUT-102 T-cell line. Only 1M3B manifested erratic binding,all the other chelating agents demonstrated similar iodine-131and yttrium-88 binding, ranging from 58 to 91% of each ofthese forms of anti-Tac bound to the Tac-expressing cells (Table2), while only 1-5% bound to Tac-negative cells (data notshown). This value was comparable to 60-90% of the 3H-labeledanti-Tac binding to the same cell line. Thus neither the conjugation with the CA nor the radiolabeling procedure altered thecapacity of the antibody to bind to its antigenic target. Thus,each of the bifunctional CAs examined fulfilled the first criterion for a suitable agent for immunotherapy.

Effect of CA Coupling Procedures on the Rates of Catabolism.Plasma clearance studies were performed in athymic nude miceto assess whether the CA coupling procedures used to covalentlyattach the various bifunctional CAs to anti-Tac altered thecatabolism rates or tissue distribution of the antibody. Specifically, the metabolism of I3'l-labeled anti-Tac-CA was comparedto the metabolism of radioiodinated unmodified anti-Tac, andthe CA-coupled antibody was examined for evidence of damageto the MAb during this initial chemical conjugation. As shownin Fig. 1 and Table 3, each of the I31l-labeled, CA-coupled

antibodies behaved in a very similar fashion to radioiodinated,unmodified anti-Tac. They showed a comparable clearanceduring the first 3 days postinjection, representing the combinedeffects of catabolism and distribution from the intravascular tothe extravascular, extracellular spaces. Furthermore, followingthis initial phase that is dominated by distribution into extra-vascular spaces, each of the '"I-labeled antibodies exhibited acomparable, prolonged terminal half-life dominated by catabolism of approximately 3-6 days, a value that agrees withprevious reports for mouse IgG ( 16,17). Therefore, the differentCAs did not alter the rate of metabolic clearance of anti-Tacwhen coupled to it, thus fulfilling the second criterion for auseful CA for immunotherapy.

Rates of Elution of Yttrium-88 from Anti-Tac-CA-^Y. Todetermine the rates of yttrium-88 elution from the chelatedanti-Tac molecules and thereby define the capacity of the different CAs to retain yttrium in vivo, we compared the in vivo

2641

on May 22, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CHELATING AGENTS FOR YTTRIUM RADIOIMMUNOTHERAPY

100.806040

UNMODIFIED

SCN-BZ-DTPA

1B-EDTA i CaDTPA

1M3B MX

2B-DTPA

1B3M

12345 12345 12345 12345

DAYS POSTINJECTIONFig. 1. Plasma clearance of ' ' ' 1(O)- and "Y (^-labeled anti-Tac-CA. Aihymie

nude mice were injected in the tail vein with 1 /¡Ciof each radionuclide-labeled,CA-coupled anti-Tac. Plasma clearance of anti-Tac-CA-"Y using CAs SCN-BZ-DTPA, IM3B, MX, and 1B3M was similar to "'I-labeled unmodified anti-Tacor I3ll-labeled anti-Tac-CA, indicating minimal elution of 88Yfrom the complex.In contrast, plasma clearance of anti-Tac-CA-MY using 1B-EDTA, CaDTPA, and2B-DTPA was much faster than their '"I-labeled counterparts, indicating yt-trium-88 was rapidly eluted from the complex.

Table 3 Analysis of blood clearance dataAnalysis of blood clearance data is described in "Materials and Methods."

CANoneCa-DTPA1B-EDTA2B-DTPARadionuclideIodine-131Iodine-131

Yttrium-88Iodine-131

Yttrium-88Iodine-131

Yttrium-88FCR"0.3440.143

0.7510.414

3.2890.398

0.926FER*0.602.8750.528

SCN-BZ-DTPA Yttrium-88 0.297

1M3B-DTPAMX-DTPA1B3M-DTPAIodine-131Yttrium-88Iodine-131

Yttrium-88Iodine-131

Yttrium-880.327

0.2070.422

0.3300.385

0.352-0.127-0.092-0.033" FCR (fractional catabolic rate) estimated from slopes and intercepts of

multiexponential curves.* FER (fractional elution rate) of Yttrium-88 is the difference between the

apparent FCR with Yttrium-88 and the FCR with Iodine-131 antibody. It reflectsthe rate of elution of yttrium from the chelate-antibody.

metabolism of anti-Tac-CA-88Y with its iodinated anti-Tac-CA

counterpart given simultaneously to the same animal. The timecourses of the clearance of different forms of anti-Tac-CA-88Yare compared with the clearance rates for l31I-labeled anti-Tac-CA in Fig. 1. Accelerated plasma clearance rates for yttrium-88 were observed with the 1B-EDTA, CaDTPA, and 2B-DTPAelídalesof anti-Tac when compared with their I3ll-labeled anti-Tac-CA counterparts. These yielded FER (FCR-88Y minusFCR-U1I) of 2.87, 0.60, and 0.53 of the intravascular pool per

day (Table 3). These rates of elution are quite high, exceedingthe rates of clearance of iodinated anti-Tac-CA in each caseand extending to as much as seven times the metabolic rate ofthe MAb alone. In contrast, anti-Tac-CA-88Y with the SCN-BZ-DTPA, 1M3B, MX, and 1B3M CAs had plasma survivalsand FCR that were equal to their I31l-labeled, CA-coupledcounterparts or I3'l-labeled, unmodified anti-Tac. With each of

these agents, no demonstrable elution of yttrium-88 from theanti-Tac-CA was observed (Fig. 1, Table 3). One can concludefrom a comparison of the iodine-131 and yttrium-88 plasmaactivity that the chelate linkage and radiolabeling proceduresper se do not alter the metabolism of the MAb. However, thecapacity of the various CAs to securely bind the yttrium-88varied markedly among the different CAs examined. Only SCN-BZ-DTPA, 1M3B, MX, and 1B3M fulfilled this importantcriterion for a useful chelating agent for radioimmunotherapy.

Organ Distribution of Yttrium-88. The primary form of tox-icity observed in yttrium-90 therapy is suppression of hemato-poiesis, which is a reflection of the deposition of the free yttriumin its target organ, the bone (18, 19). We determined thedistribution of radiolabeled anti-Tac-CA on Day 5 following

i.v. administration for each of the CAs in the study. Simultaneous studies of the amount of yttrium-88 and iodine-131 weremade from the blood and each of the organs using '" I-labeledanti-Tac-CA as an approximation of plasma activity in orderto correct for the yttrium-88 retained in the blood. Yttrium-88exhibited only modest accumulation in the eight tissues examined, slightly elevated yttrium to iodine ratios were seen in liver,spleen, and kidney for all chelates (Table 4). This is due to thedifference in the end metabolic products wherein iodine rapidlydiffuses from the organ while yttrium is trapped in this site(20). The exception is in different levels of yttrium-88 boneaccumulation (Fig. 2). Following subtraction of I31l-plasmaactivity, there was only 1.4-1.8% of the injected dose per gram

of bone, for those CAs (1M3B, MX, and 1B3M) that had along plasma survival of the M Ab-CA-88Y (Fig. 2); that is, there

was very little accumulation of the radionuclide in bone whenthere was no demonstrable elution of yttrium from the conjugate. In contrast, for the three CAs (CaDTPA, 2B-DTPA, and1B-EDTA) that had manifested rapid clearance from theplasma, yttrium-88 exhibited much greater accumulation in thebone, with a range from 5-14% of the injected dose per gramof femur (Fig. 2). Thus, for these CAs the increased depositionin the bone paralleled the rapid clearance from the plasma andelution of yttrium-88 from the conjugate. This increased deposition in bone, the target organ of yttrium toxicity, when usingthe unstable CAs is emphasized by the ratio of yttrium to iodineactivity in this organ. This ratio in the bone was 3.6 with thestable lM3B-chelating agent, whereas it was 86.8 with theunstable CaDTPA-chelating agent. These studies clearly dem-

Table 4 Labeled CA-coupled antibody

Organ distribution (Day 5) for representative stable and unstable yttriumchelates. Femurs and bone marrow were separated, and organs were rinsed andblotted dry before weighing. Data are expressed as percentage injected dose pergram of tissue. Values are not corrected for radioactivity present in the blood ofthe organ.

CA1M3B-DTPACa-DTPATissueHeartLungsLiverSpleenKidneysIntestineFemurMarrowHeartLungsLiverSpleenKidneysIntestineFemurMarrowYttrium-882.97

±0.425.52±0.835.57

±0.515.84±1.414.64±0.821.66

±0.272.53±0.291.13

±0.201.41

±0.282.70±0.346.25±1.297.11

±4.044.16±1.111.07

+0.2912.15+1.341.50

+ 0.80Yttrium-88/

Iodine-131Iodine-1311.84

+0.332.90±0.542.18

±0.452.02±0.822.28

±0.580.97±0.220.71

±0.150.52+0.112.61

+0.403.83±0.022.92+0.233.53

±0.582.64+0.471.36+0.190.14

±0.030.98+ 0.211.61.92.62.92.01.73.62.20.50.72.12.01.60.886.81.5

2642

on May 22, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CHELATING AGENTS FOR YTTRIUM RADIOIMMUNOTHERAPY

1M3B-DTPA

MX-DTPA

1B3M-DTPA

Ca-DTPA

2B-DTPA

1B-EDTA

_L8 10 12 14

ID/GRAM TISSUE

16 18

Fig. 2. Yttrium and iodine bone activity. Animals were injected with 13'I-labeled anti-Tac-CA and anti-Tac-CA-"Y. The fifth day postinjection animals

were sacrificed, and femurs were flushed of their bone marrow before beingcounted. CAs for which plasma survival of yttrium XXwas prolonged, had lowbone uptake of yttrium-88 expressed as percent of the injected dose (%ID) pergram of tissue. Animals were the same as in Fig. 1. D, iodine-131 activity; D.yttrium-88 activity retained in bone.

onstrate that the 1B-EDTA, CaDTPA, and 2B-DTPA CAs arepoor choices for radioimmunotherapy, with yttrium-90, sincethey do not stably complex yttrium-88, thus leading to itsdeposition in the bone.

DISCUSSION

Bifunctional CAs are essential components in the assemblyof radiometal-labeled MAbs and play a critical role in determining the in vivo stability of the radioconjugate and, thus, itseffectiveness in irnmunotherapy. During the past 5 years, attention has focused on the polyaminocarboxylic acids EDTA andDTPA as CAs for radioimaging and radioimmunotherapy. Inattempts to improve the stability of these CAs, derivatives havebeen synthesized and side arms have been added to them inorder to render them heterobifunctional and to maintain theirfull denticity (number of metal binding sites). However, manyof these CAs still result in inadequate targeting. We modifiedthe benzyl DTPA backbone by adding methyl substituents inhopes of decreasing the off rate of yttrium from the chelate-coupled antibody. As noted above, CAs suitable and selectedfor radioimmunotherapy should fulfill the following criteria:(a) the CA coupled to the MAb should not compromise antibody immunospecificity, (/>) the chelation and radiolabelingprocedure should not alter the metabolism of the MAb, (c) theCA should not prematurely release the radiometal in vivo, and(d) the CA should help clear the radiometal following catabo-lism. A number of the chelates that have been widely used forradioimmunoimaging and radioimmunotherapy do not fulfillthe third criterion, thus allowing for the elution of the radiometal from the conjugate in vivo. This elution results in areduction in the effectiveness of therapy as well as a markedincrease in toxicity. Furthermore, the use of such CAs can leadto the unwarranted conclusion that certain radionuclides arenot satisfactory for irnmunotherapy because of undesirable toxicity.

Yttrium-90 is considered a favorable radiometal for irnmunotherapy because of its intense ßemission (>99%) with an£maxvalue of 2.3 Mev; in addition, it lacks -y-ray emissions, hasa 64-h half-life, and decays to a stable daughter. Yttrium-90 isalso conveniently available in large quantities (~100 mCi/batch) from a strontium-90 generator (21). Yttrium-88 was

used in the present analysis of in vivo survival, elution, anddistribution because it has the same chemical behavior as yttrium-90 but, in contrast to yttrium-90, can be accurately quantified because of its strong 7-ray emission.

When yttrium is released from its CA, it has been shown tolocalize to bone (18,19), causing potentially dose-limiting bonemarrow toxicity. Studies using the widely tested cyclic anhydride DTPA-chelating agent for antibody linkage have reportedyttrium accumulation in the liver (22-24) and bone (24), withAnderson-Berg, Squire, and Strand demonstrating bone accumulation equal to that observed in animals given 90Y-acetate

(25). Despite the evidence that CaDTPA is unstable in vivo andthe call for the better reagents, there are still attempts to useCaDTPA in therapeutic trials. Recently, Hnatowich and co-workers (26) reported very high doses of yttrium-90 in the bone(as high as 50% of the injected dose) following i.p. administration of a MAb-CaDTPA-90Y complex to patients with ovarian

cancer. It is clear from these studies, as well as our own, thatCaDTPA, 1B-EDTA, and 2B-DTPA are not appropriate CAsfor irnmunotherapy with yttrium-90.

However, we have demonstrated that four backbone-substituted derivatives of DTPA (SCN-BZ-DTPA, 1M3B, MX, and1B3M) fulfill the three criteria for suitable yttrium-labeledradioimmunotherapeutic agents. Chelation and radiolabelingprocedures did not compromise antibody specificity. In addition, these CAs and the procedures required to couple them tothe MAb did not alter the metabolism of the antibody conjugates. Furthermore, as demonstrated by the parallel 88Y- and131I-labeled anti-Tac plasma clearance studies, there was no

measurable elution of the yttrium from these four conjugates.The methyl substituents were added based first on order of easeof synthesis so subtle differences were observed but were notanticipated to vary widely. These observations contrast withthose resulting from studies of CaDTPA, 1B-EDTA, and 2B-DTPA, which displayed rapid blood clearance of yttrium. Furthermore, the CAs that lost the yttrium rapidly were associatedwith a high level (5-15%) of the yttrium injected dose per gramof tissue appearing in the bone, whereas the stable CAs had lessthan 2% injected dose per gram bone accumulation. In thesestudies we have assumed but have not proved that yttrium boneactivity is due to the deposition of free yttrium rather than acomplex.

We conclude that the use of the stable chelating agents, SCN-BZ-DTPA, 1M3B, MX, and 1B3M, should minimize toxicityto bone, bone marrow, and other normal tissues, while maximizing the dose delivered to the target tissue. This maximization of the therapeutic/toxicity ratios should lead to the realization of the full potential of MAb-CA-radionuclides as radioimmunotherapeutic agents.

REFERENCES

1. Catane, R., and Longo, D. L. Monoclonal antibodies for cancer therapy. Isr.J. Med. Sci., in press, 1989.

2. Eckelman, W. C., Paik, C. H., and Reba, R. C. Radiolabeling of antibodies.Cancer Res., 40: 3036-3042, 1980.

3. Wessels, B. W., and Rogers, R. D. Radionuclide selection and model absorbeddose calculation for radiolabeled tumor associated antibodies. Med. Phys.,//: 638-645, 1984.

4. Meares, C. F., and Goodwin, D. A. Linking radiometals to proteins with

2643

on May 22, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

CHELATING AGENTS FOR YTTRIUM RADIOIMMUNOTHERAPY

bifunctional chelating agents. J. Prot. Chem., 4: 215-228, 1984.5. Keenan, A. M.. Harbert, J. C"..and Larson, S. M. Monoclonal antibodies in

nuclear medicine. J. Nucí.Med., 26: 531-537, 1985.6. Waldmann, T. A., Leonard, W. J., Depper, J. M., Kronke, M., Thompson,

C. B., Kozak, R., and Greene, W. C. Structure, function, and expression ofthe receptor for interleukin-2 on normal and malignant lymphocytes. CancerCell, 3: 221-226, 1985.

7. Margerum, D. W., Cayley, G. R., Weatherburn, D. C., and Pagenkopf, G.K. Kinetics and mechanisms of complex formation and ligand exchange. In:A. E. Marteli (ed.), Coordination Chemistry, Monograph 174, Vol. 2, Chapter 1, pp. 1-220. Washington, DC: American Chemical Society, 1978.

8. Leonard, W. J., Depper, J. M., Uchiyama, T., Smith, K. A., Waldmann, T.A., and Greene, W. C. A monoclonal antibody that appears to recognize thereceptor for human T cell growth factor. Nature (Lond.), 300: 267-269,1982.

9. Brechbiel, M. W., Gansow, O. A., Atcher, R. W., Schlom, J., Estaban, J.,Simpson, D. F.. and Colcher, D. Synthesis of l-(/>-isolhiocyanatobenzyl)derivatives of DTPA and EDTA. Antibody labeling and tumor-imagingstudies. Inorganic Chemistry, 25: 2772-2781, 1986.

10. Brechbiel, M. W. New bifunctional chelates for radioimmunoimaging andtherapy. Ph.D. Thesis, The American University, 1988.

11. McFarlane, A. S. Metabolism of plasma proteins. In: H. N. Munro and J. B.Allison (eels.). Mammalian Protein Metabolism, Vol. 1, Chap. 8, pp. 331-336. New York: Academic Press, 1984.

12. Waldmann, T. A. Methods for the study of the metabolism of immunoglob-ulins. Methods Enzymol., 116: 201-210, 1985.

13. Tack, B. F., Dean, J., Eilat, D., Lorenz, P. E., and Scheeler, A. N. Tritiumlabeling of proteins to high specific radioactivity by reductive methylation. J.Biol. Chem., 255: 8842-8847, 1980.

14. Waldmann, T. A., and Strober, W. Metabolism of immunoglobulins. Prog.Allergy, 13: 1-110, 1969.

15. Matthews, C. M. E. The theory of tracer experiments with 13'I-labelledplasma proteins. Phys. Med. Biol., 2: 36-53, 1957.

16. Fahey, J. 1... and Sell, S. The immunoglobulins of mice. V. The metabolicproperties of five immunoglobulin classes. J. Exp. Med., 122:41-58, 1965.

17. Humphrey, J. H., and Fahey, J. L. The metabolism of normal plasma proteinsand gamma-myeloma protein in mice bearing plasma-cell tumors. J. Clin.Invest., 40:1696-1705, 1961.

18. Durbin, P. W. Metabolic characteristics within a chemical family. HealthPhysics, 2: 225-238, 1960.

19. O'Mara, R. E., McAfee, .1.G., and Subramanian, G. Rare earth nuclides aspotential agents for skeletal imaging. J. Nucí.Med., 10: 49-51, 1968.

20. Pimm, M. V., Perkins, A. C., and Baldwin, R. W. Differences in tumor andnormal tissue concentrations of iodine- and indium-labelled monoclonalantibody. Eur. J. Nucí.Med., ;;.•300-304, 1985.

21. Chinaol, M., and Hnatowich, D. J. Generator-produced yttrium-90 forradioimmunotherapy. J. Nucí.Med., 28:1465-1470, 1978.

22. Buchsbaum, D. J., Hann, D. E., Randall, B. C., Buchegger, F., and Mach, J.P. Radiolabelling of monoclonal antibody against carcinoembryonic antigenwith "Y and biodistribution studies. Int. J. Nucí.Med. Biol., 12: 79-82,

1985.23. Hnatowich, D. J., Virzi, F., and Doherty, P. W. DTPA-coupled antibodies

labelled with yttrium-90. J. Nucí.Med., 26: 503-509, 1985.24. Vaughan, A. T. M., Keeling, A., and Ynakuba, S. C. S. The production and

biological distribution of yttrium-90 labelled antibodies. Int. J. Appi. Radiât.Isot., 36:803-806, 1985.

25. Anderson-Berg, W. T., Squire, R. A., and Strand, M. Specific radioimmunotherapy using 90Y-labelled monoclonal antibody in erythroleukemic mice.Cancer Res., 47:1905-1912,1987.

26. Hnatowich, D. J., Chinol, M., Siebecker, D. A., Gionet, M., Griffin, T.,Doherty, P. W., Hunter, R., and Käse,K. R. Patient biodistribution ofintraperitoneally administered yttrium-90-labeied antibody. J. Nucí.Med.,29:1428-1434, 1988.

2644

on May 22, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

1989;49:2639-2644. Cancer Res   Robert W. Kozak, Andrew Raubitschek, Saed Mirzadeh, et al.   Toxicity

Survival and Organin VivoCritical Factors in Determining Radioimmunotherapy with Yttrium-90 Monoclonal Antibodies: Nature of the Bifunctional Chelating Agent Used for

  Updated version

  http://cancerres.aacrjournals.org/content/49/10/2639

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/49/10/2639To request permission to re-use all or part of this article, use this link

on May 22, 2020. © 1989 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from