Preparation and in vivo evaluation of linkers for 211At labeling of humanized anti-Tac.

The syntheses, radiolabeling, antibody conjugation, and in vivo evaluation of new linkers for 211At labeling of humanized anti-Tac (Hu-anti-Tac), an antibody to the alpha-chain of the IL-2 receptor (IL-2Ralpha) shown to be a useful target for radioimmunotherapy are described. Synthesis of the organometallic linker precursors is accomplished by reaction of the corresponding bromo- or iodoaryl esters with bis(tributyltin) in the presence of a palladium catalyst. Subsequent conversion to the corresponding N-succinimidyl ester and labeling with 211At of two new linkers, N-succinimidyl 4-[211At]astato-3-methylbenzoate and N-succinimidyl N-(4-[211At]astatophenethyl)succinamate (SAPS), together with the previously reported N-succinimidyl 4-[211At]astatobenzoate and N-succinimidyl 3-[211At]astato-4-methylbenzoate, are each conjugated to Hu-anti-Tac. The plasma survival times of these conjugates are compared to those of directly iodinated (125I) Hu-anti-Tac. The N-succinimidyl N-(4-[211At]astatophenethyl)succinamate compound (SAPS) emerged from this assay as the most viable candidate for 211At-labeling of Hu-anti-Tac. SAPS, along with the directly analogous radio-iodinated reagent, N-succinimidyl N-(4-[125I]astatophenethyl)succinamate (SIPS), are evaluated in a biodistribution study along with directly iodinated (125I) Hu-anti-Tac. Blood clearance and biological accretion results indicate that SAPS is a viable candidate for further evaluation for radioimmunotherapy of cancer.

In vivo evaluation of bismuth-labeled monoclonal antibody comparing DTPA-derived bifunctional chelates.

Among the radionuclides considered for radioimmunotherapy, alpha-emitters such as the bismuth isotopes, 212Bi and 213Bi, are of particular interest. The macrocyclic ligand, DOTA, has been shown to form stable complexes with bismuth isotopes. The kinetics of the complexation of bismuth with the DOTA chelate, however, are slow and impractical for use with 212Bi and 213Bi that have half-lives of 60.6 and 45.6 min. The study described herein compares six DTPA derived bifunctional chelates with the goal of identifying an alternative to the DOTA ligand for radiolabeling with bismuth. Radioimmunoconjugates comprised of MAb B72.3, each of the six DTPA chelates, and radiolabeled with 206Bi, which facilitated the evaluation due to its readily detectable gamma-emission. In vitro studies showed that each of the radioimmunoconjugates retained immunoreactivity that was comparable to its 125I-labeled counterpart. The 206Bi- and 125I-labeled immunoconjugates were then co-injected i.p. into normal athymic mice. Injection of Afree@ 206Bi demonstrated that the kidneys were the critical organ to evaluate for retention of bismuth in the chelate complex. Major differences were identified among the six preparations. The CHX-A and -B immunoconjugates were found to have 1) the lowest %ID/gm in the kidney; 2) a level of 206Bi in the kidney that was comparable to that of 125I-B72.3; and 3) no significant uptake of 206Bi evident in other organs such as bone, lung and spleen. The results described herein suggest that either of the cyclohexyl derivatives of DTPA may be suitable candidates for the labeling of immunoconjugates with alpha-emitting bismuth isotopes for radioimmunotherapeutic applications.

Alpha-emitting bismuth cyclohexylbenzyl DTPA constructs of recombinant humanized anti-CD33 antibodies: pharmacokinetics, bioactivity, toxicity and chemistry.

UNLABELLED: The alpha-particle-emitting radionuclides have several physical characteristics that make them attractive candidates for radioimmunotherapy: (a) high linear energy transfer; (b) short path lengths (50-80 microm); and (c) limited ability of cells to repair damage to DNA. This article describes the pharmacokinetic, bioactivity, toxicity and chemical characteristics of alpha-particle-emitting, 213Bi and 212Bi radiometal conjugated HuM195 (anti-CD33) constructs. Conjugation of HuM195 to SCN-CHX-A-DTPA resulted in the attachment of up to 10 chelating ligand molecules per antibody. RESULTS: Radiolabeling efficiency of the CHX-A-DTPA-HuM195 construct with 213Bi was 78%+/-10% (n = 46) after 10 min at specific activities of up to 1110 MBq/mg. The immunoreactivity of the 213Bi-labeled CHX-A-DTPA-HuM195 construct was 84%+/-10% (n = 28) and was independent of the specific activity. The bismuth-labeled CHX-A-DTPA-HuM195 construct was rapidly internalized into the cell in a time-dependent manner ranging from 50% at 1 h to 65% at 24 h. 205Bi/206Bi-labeled constructs were stable for at least 2 d in vitro in the presence of human serum at 37 degrees C. After injection into mice, there was no uptake or loss of bismuth to mouse tissues, which do not express CD33, or to the kidney, which has avidity for free bismuth. Mice injected intraperitoneally with doses of (213Bi)CHX-A-DTPA-HuM1 95 ranging from 18.5 to 740 MBq/kg showed no toxicity, but at 2590 MBq/kg, two of the three mice died within 2 wk and a third mouse showed significant reductions in white blood cell counts. Mice injected intravenously with doses of (213Bi)CHX-A-DTPA-HuM195 up to 370 MBq/kg exhibited little toxicity, but 666 MBq/kg was above the MTD for mice. Leukemia cell killing in vitro with bismuth-labeled HuM1 95 showed dose- and specific activity-dependent killing of CD33+ HL60 cells; approximately 50% killing was observed when two bismuth atoms (50 fM radiolabeled antibody) were initially bound onto the target cell surface. CONCLUSION: Alpha-emitting antibodies are among the most potent cytotoxic agents known, yet are specific and appear safe in vivo. The physical and biochemical characteristics of the 213Bi isotope and its generation, as well as the biochemistry of the 213Bi-labeled CHX-A-DTPA-HuM195 construct, make it possible to use the constructs safely and feasibly in humans at therapeutic levels.

Similarities and differences in 111In- and 90Y-labeled 1B4M-DTPA antiTac monoclonal antibody distribution.

UNLABELLED: Monoclonal antibodies (MoAb) labeled with 90Y are being used for radioimmunotherapy. Because 90Y is a beta emitter, quantitative information from imaging is suboptimal. With the concept of a "matched pair" of isotopes, 111In is used as a surrogate markerfor90Y. We evaluated the differences in biodistribution between 111In- and 90Y-labeled murine antiTac MoAb directed against the IL-2Ralpha receptor. METHODS: The antiTac was conjugated to the 2-(4-isothiocyanatobenzyl)-6-methyl-diethylenetriamine pentaacetic acid (1B4M-DTPA, also known as MX-DTPA). Nine patients with adult T-cell leukemia were treated. Patients received approximately 185 MBq (5 mCi) 111In-labeled antiTac for imaging and 185-555 MBq (5-15 mCi) 90Y-labeled antiTac for therapy. The immunoreactivity of 111In-labeled antiTac was 90%+/-6%, whereas for 90Y-labeled antiTac, it was 74%+/-12%. RESULTS: The differences in blood and plasma kinetics of the two isotopes were small. The area undemeath the blood radioactivity curve was 1.91 percentage+/-0.58 percentage injected dose (%ID) x h/mL for 111In and 1.86%+/-0.64 %ID x h/mL for 90Y. Urinary excretion of 90Y was significantly greater than that of 111In in the first 24 h (P = 0.001), but later, the excretion of 111In was significantly greater (P = 0.001 to P = 0.04). Core biopsies of bone marrow showed a mean of 0.0029+/-0.0012 %ID/g for 111In, whereas the 90Y concentration was 0.0049+/-0.0021 %ID/g. Analyses of activity bound to circulating cells showed concentrations of 500-30,000 molecules of antiTac per cell. When cell-bound activity was corrected for immunoreactive fraction, the ratio of 111In to 90Y in circulating cells was 1.11+/-0.17. Three biopsies of tumor-involved skin showed ratios of 111In to 90Y of 0.7, 0.9 and 1.1. CONCLUSION: This study shows that differences typically ranging from 10% to 15% exist in the biodistribution between 111In- and 90Y-labeled antiTac. Thus, it appears that 111In can be used as a surrogate marker for 90Y when labeling antiTac with the 1 B4M chelate, although underestimates of the bone marrow radiation dose should be anticipated.

In vivo comparison of CHX-DTPA ligand isomers in athymic mice bearing carcinoma xenografts.

Monoclonal antibodies (MAbs) labeled with radiometallonuclides via metal chelators are being investigated in the laboratory for use in clinical trials. The biodistribution of 111In- and 88Y-labeled antibody (MAb B72.3) using two isomeric forms (CHX-A and CHX-B) of the 2-(p-isothiocyanatobenzyl)-cyclohexyl-DTPA was compared in athymic mice bearing LS-174T tumors, human colon carcinoma xenografts. CHX-(A or B)-125I-DTPA-B72.3 was co-injected in all athymic mice to assess if the chelate conjugation altered the properties of MAb B72.3. In vitro studies demonstrated maintenance of integrity and immunoreactivity for both radioimmunoconjugates. The in vivo analysis, however, indicated major differences between the two isomer forms. In fact, the 88Y-CHX-A-DTPA radioimmunoconjugate demonstrated over the 7-day study period, a more efficient and stable tumor localization as well as a slower blood clearance rate than the CHX-B-DTPA chelate conjugate, suggesting a greater in vivo stability. Differences were also evident in critical normal organ uptake: no significant increase in liver- and spleen- or bone-to-blood ratios was observed when the CHX-A-DTPA chelate was labeled with indium or yttrium. The results described here demonstrate that the CHX-A-DTPA chelate conjugate can be considered more suitable than the CHX-B-DTPA isomer form when radiometallonuclides are coupled to an MAb.

Physical parameters and biological stability of yttrium(III) diethylenetriaminepentaacetic acid derivative conjugates.

The solution equilibria, acid dissociation, and serum stability of a series of Y(III) complexes of DTPA ligands functionalized with p-nitrobenzyl, methyl, and trans-cyclohexyl substituents were studied. The thermodynamic stability of the complexes studied ranged from log K = 21.53 to 24.7. Acid dissociation rates were found to decrease as the substitution on the carbon backbone increased, and significant differences in dissociation rates were observed for the Y(III) complexes of a pair of diasteriomeric cyclohexyl-DTPA ligands. While one diastereomer was found to have the slowest acid dissociation rate of the entire DTPA series, it was remarkably labile in both serum stability and in vivo studies.

In vivo evaluation of a lead-labeled monoclonal antibody using the DOTA ligand.

The aim of this study was to assess the utility of a radioimmunoconjugate containing a lead radionuclide for therapy and scintigraphy applications. The radioimmunoconjugate evaluated consisted of a bifunctional DOTA ligand and monoclonal antibody (MAb) B72.3 using athymic mice bearing LS-174T tumors, human colon carcinoma xenografts. In the studies reported here, the lead-203-DOTA complex itself was first demonstrated to have in vivo stability. MAb B72.3 was then conjugated with the DOTA ligand and labeled with 203Pb, and the immunoreactivity of B72.3 was maintained. The localization of the radioimmunoconjugate to tumor tissue and other select organs paralleled that of DOTA-125I-B72.3, suggesting a similar metabolic pattern of the two radioimmunoconjugates. Thus, the DOTA-metal complex does not alter the behavior of the radioimmunoconjugate. Tumor localization of the 203Pb-DOTA-B72.3 conjugate was demonstrated with biodistribution studies as well as immunoscintigraphy studies. Such data highlight the stability of a lead radionuclide in the DOTA ligand. The suitability of this chelation chemistry for labeling radioimmunoconjugates with a lead radionuclide now makes its application in nuclear medicine a feasible proposition.

Evaluation of the in vivo biodistribution of yttrium-labeled isomers of CHX-DTPA-conjugated monoclonal antibodies.

UNLABELLED: We evaluated the in vivo stability and biodistribution of four isomers (CHX-A', CHA-A", CHX-B' and CHX-B") of 2-(p-isothiocyanatobenzyl)-cyclohexyl-diethylenetriaminepentaaceti c acid (CHX-DTPA), a recently developed backbone-substituted derivative of DTPA. METHODS: The ligands were conjugated to monoclonal antibody B3, a murine IgG1 kappa, and labeled with 88Y at 55.5-66.6 MBq/mg (1.5-1.8 mCi/mg). Nontumor-bearing nude mice were injected intravenously with 55.5-66.6 kBq (1.5-1.8 microCi) of 88Y-labeled B3 conjugates and with 125I-labeled B3 as an internal control. The mice were then killed at 6, 24, 48, 96 and 168 hr postinjection. RESULTS: At 168 hr, the concentration of 88Y in processed bone of either CHX-A' [4.6% injected dose (ID)/g] or CHX-A" (4.0%ID/g) was less than that of either the CHX-B' (21.9%ID/g) or B" (12.1%ID/g) ligands. The two ligands CHX-B" and CHX-B' were not acceptable for yttrium labeling of antibody because of their high and progressive bone accumulation. The accumulation of 88Y in bone of CHX-B' was five times greater than that of CHX-A' at 168 hr. The CHX-A" cleared from the circulation slightly faster than CHX-A' without releasing the yttrium and showed the lowest uptake by bone of any of the four isomers. The accumulation in the other normal organs was similar for all four isomers of 88Y-CHX-B3 conjugates. CONCLUSION: Although the CHX-B" and CHX-B' were not acceptable for labeling with yttrium, the CHX-A' and CHX-A" were suitable, indicating that differences in stereochemistry can greatly influence stability of radionuclide in the chelate.

Preparation of 211At-labeled humanized anti-Tac using 211At produced in disposable internal and external bismuth targets.

These studies describe the production and purification of 211At as well as the procedure for labeling humanized anti-Tac, the antibody to the alpha-chain of the IL-2 receptor (IL-2R alpha), which has been shown to be a useful target for immunotherapy. The optimized protocol combines the advantages of the two-stage dry distillation procedure with the astatination of trialkylstannyl substances as labeling compounds for proteins. The 211At was produced by bombarding either an external or a recently developed disposable internal bismuth target with alpha-particles from a Cyclotron Corporation CS-30 cyclotron. The 211At was found to contain less than 0.01% 210At. The production rate for the external target was 0.15 mCi +/- 0.056 microA(-1) h(-1) (n = 9) (5.55 MBq mcroA[-1] h[-1]). The production rate for the internal target was 0.44 +/- 0.14 mCi microA(-1) h(-1) (n = 16) (16.28 MBq mcroA[-1] h[-1]).

Rapid and specific targeting of monoclonal antibody A33 to a colon cancer xenograft in nude mice.

Monoclonal antibody (mAb) A33 detects a glycoprotein homogeneously expressed by > 95% of human colon cancers and by normal colon cells. The A33 antigen is not secreted or shed and after mAb A33 binds to antigen on the cell membrane, a fraction of membrane-bound mAb A33 is internalized into endosomes. Phase I 131I-mAb A33 biodistribution studies have shown consistent, specific tumor-targeting, and phase I radioimmunotherapy trials with 131I- or 125I-mAb A33 have demonstrated antitumor effects. Here we describe a nude mouse model that was established using a human colon cancer cell line, SW1222, which grows as a relatively hypovascular, invasive heterotransplant when injected i.m. Peak uptake of 131I-labeled or 111In-chelated mAb A33 was observed at 48-96 h, with a mean of 34% (SE +/- 5.0) and 46.7% (SE +/- 1.7) injected dose per gram of tumor tissue, respectively. 111In-mAb A33 was retained in tumor tissue longer than halide radioimmunoconjugates. The specificity of antibody localization was assessed using a control antibody (tumor uptake and pharmacokinetics), a control tumor, corrections for vascular antibody blood-pooling in tumor tissue, and blocking of radiolabeled mAb A33 localization by pretreating mice with excess unlabeled mAb A33. These experiments demonstrate that mAb A33 localization in tumor was specific, and they emphasize the unexpected rapidity with which the antibody localizes. Our conclusions were confirmed by immunohistochemical techniques which allowed direct visualization of localization and distribution of the humanized version of mAb A33 in tumor tissue. Furthermore, antibody doses approximating tumor-saturating doses demonstrated that a homogeneous distribution of antibody in tumor is possible. This model will be valuable for studies focusing on general physiologic aspects of antibody-to-tumor cell localization and critical as a guide to the evaluation of various A33 antibody constructs and combinations with other therapies for the treatment of colon cancer.

Stereochemical influence on the stability of radio-metal complexes in vivo. Synthesis and evaluation of the four stereoisomers of 2-(p-nitrobenzyl)-trans-CyDTPA.

Distinct differences in in vivo stability of the two diastereomeric C-Functionalized CyDTPA chelating agents, (CHX-A DTPA and CHX-B DTPA, both racemates), as recently reported prompted further investigation as to why differences in configuration produced striking effects on the in vivo stability of their yttrium complexes. To this end, the four individual component stereoisomers of CHX-A and CHX-B were synthesized and ability to bind yttrium was investigated both in vitro and in vivo.

Factors influencing the in vivo pharmacokinetics of peptides and antibody fragments: the pharmacokinetics of two PET-labeled low molecular weight proteins.

Monoclonal antibodies (MoAbs) were proposed as candidates for selective tumor targeting based on their high binding affinity for tumors and the absence of binding by normal tissue. However, the exclusive and complete transport of the drug have been found lacking in the use of intact MoAbs, especially in the case of solid tumors. Smaller fragments that maintained the desiderable tumor targeting characteristics appear to have an advantage because of the increase in whole body clearance and the shorter time to maximum target to non-target ratio. But the binding to normal tissue increases as the molecular weight or size decreases, especially in the kidney, because the glomerular sieving coefficient increases. We have used two approaches to overcome the normal tissue binding: a) the use of lysine to block the uptake of the Low Molecular Weight Proteins (LMWP) in the proximal tubules and b) the labeling of different amino acids in the LMWP to alter the residence time in the kidney. The combination of these two methods, blocking the uptake with lysine and shortening the residence time of the radioactive compound produced in the kidney, can lead to substantially decreased normal kidney targeting. The lysine paradigm was effectively demonstrated using 18F-labeled anti-Tac dsFv. Shortening the residence time was illustred by comparing the kinetics of the lysine catabolite versus the methionine metabolite. Furthermore, the rapid pharmacokinetics of the LMWP are consistent with the use of the shorter half-lives of PET radionuclides and the concomitant increase in quantitation and sensitivity afforded by PET.

Purification of p-nitrobenzyl C-functionalized diethylenetriamine pentaacetic acids for clinical applications using anion-exchange chromatography.

A general process for the purification of large quantities (5-10 g) of p-nitrobenzyl C-functionalized diethylenetriamine pentaacetic acids is reported. The method of choice to achieve purification for clinical applications is anion-exchange chromatography.

In vitro and in vivo evaluation of structure-stability relationship of 111In- and 67Ga-labeled antibody via 1B4M or C-NOTA chelates.

2-(p-isothiocyanatobenzyl)-1,4,7-triazacyclononane-1,4,7-triacetic acid (C-NOTA) or 2-(p-isothiocyanatobenzyl)-6-methyl-diethylenetriamine pentaacetic acid (1B4M) was conjugated to monoclonal antibody T101 (IgG2a), radiolabeled with 111In or 67Ga and then purified through size-exclusion HPLC. 111In 1B4M-T101 and 67Ga C-NOTA-T101 were stable in in vitro serum at 37 degrees C. In contrast, 111In C-NOTA-T101 and 67Ga 1B4M-T101 were unstable. The biodistribution in normal mice reflected the instability of the metal complex; the less-stable 111In C-NOTA conjugate left less tracer in blood, but more in liver and kidney whereas the less-stable 67Ga 1B4M conjugate left less tracer in blood, but more in bone. The biodistribution data suggest that the difference shown between the 111In and 67Ga conjugates might be mediated by differences in the in vivo chemistry of the metallic ions.

Radioimmunotherapy targeting of HER2/neu oncoprotein on ovarian tumor using lead-212-DOTA-AE1.

UNLABELLED: The specificity, toxicity and efficacy of lead (212Pb) radioimmunotherapy were evaluated in nude mice bearing the SK-OV-3 human ovarian tumor cell line expressing the HER2/neu proto-oncogene. METHODS: The therapeutic agent used was the tumor-specific anti-HER2/neu monoclonal antibody AE1 conjugated to 212Pb, 212Bi being the daughter and thus the source of the alpha-particle and beta emissions. A bifunctional derivative of tetraazacyclododecanetetraacetic acid (p-SCN-Bz-DOTA) was used to couple 212Pb to the anti-HER2/neu monoclonal antibody AE1. The chelating agent did not alter the binding affinity to its antigenic target or the pharmacokinetics and tissue distribution of the AE1 antibody. Toxicity and therapeutic efficacy of 212Pb-AE1 were evaluated in nude mouse ascites or solid tumor models, wherein SK-OV-3 cells were administered i.p. or s.c., respectively. RESULTS: The dose-limiting acute toxicity after i.v. administration of 212Pb-AE1 was bone marrow suppression, which was observed at doses above 25 microCi. Therefore, doses of 10 and 20 microCi were used in efficacy trials. The i.p. administration of 212Pb-AE1 3 days after i.p. tumor inoculation led to a significant (P2 = 0.015) prolongation of tumor-free survival. In a second model, i.v. treatment with 212Pb-AE1 3 days after s.c. tumor inoculation prevented subsequent tumor development in all animals treated with 10 or 20 microCi of 212Pb-AE1 (P2 = 0.002 compared to control groups). This efficacy in the adjuvant setting was antibody specific because treatments with equivalently labeled control antibody or unlabeled AE1 antibody or no treatment were less effective. The rate of growth of small (mean tumor volume, 15 mm3) SK-OV-3 tumors was modestly inhibited. However, tumor growth was not inhibited in mice bearing larger (mean tumor volume, 146 mm3) SK-OV-3 tumors by the administration of a single dose of 10 or 20 microCi of 212Pb-AE1. CONCLUSION: Lead-212-AE1 as an intact radiolabeled monoclonal antibody may be of only modest value in the therapy of bulky solid tumors due to the short physical half-life of 212Pb and time required to achieve a useful tumor-to-normal tissue ratio of radionuclide after administration. However, the radiolabeled monoclonal antibody may be useful in therapy of tumors in the adjuvant setting. Furthermore, 212Pb may be of value in select situations, including treatment of leukemia, intercavitary therapy or strategies that target vascular endothelial cells of tumors.

Evaluation of dithiol chelating agents as potential adjuvants for anti-IL-2 receptor lead or bismuth alpha radioimmunotherapy.

The dithiol chelating agents 2,3-dimercapto-1-propanesulfonic acid (DMPS) and meso-2,3-dimercaptosuccinic acid (DMSA) were evaluated for use as potential adjuvants to reduce or prevent radiotoxicity in anti-interleukin-2 receptor (IL-2R) Lead-212 or Bismuth-212 alpha-radioimmunotherapy. DMPS was less toxic than DMSA to tumor cell lines in culture. No adverse effects on the ability of an anti-IL-2R monoclonal antibody (MAb) to bind to its specific antigen were detected using DMPS or DMSA at concentrations up to 600 ug/mL in 10% or 100% mouse serum. After a 5-day oral administration of chelating agent, neither acute nor chronic toxicities on blood hematology, blood chemistry or organ weights were observed for treated mice. DMPS and DMSA were effective in accelerating whole body clearance of the gamma-emitting tracer Bismuth-206. Both chelates significantly reduced femur uptake of tracer when compared to nontreated control mice. However, only DMPS prevented early (2 h postinjection) renal accumulation. These studies support the use of DMPS as a potential adjuvant chelation therapy in Lead-212 or Bismuth-212 radioimmunotherapy protocols.

Comparison of the targeting characteristics of various radioimmunoconjugates for radioimmunotherapy of neuroblastoma: dosimetry calculations incorporating cross-organ beta doses.

To optimize the efficacy of radioimmunotherapy (RIT), the ideal antibody-radioisotope combinations should be used to deliver the highest tumor and the lowest normal tissue doses. In a mouse model, tumor and critical organ-absorbed doses delivered by different radioimmunoconjugates were calculated and compared. We used a Medical Internal Radiation Dosimetry (MIRD)-style mouse dosimetry model that incorporates cross-organ beta doses to make refined estimates of the radiation absorbed dose to tissues. Biodistribution data from neuroblastoma xenografted nude mice were used to estimate tumor, organ and bone marrow absorbed dose values for 90Y-3F8, 131I-3F8 and 131I-F(ab')2 fragments. Immunoreactive fractions of the radiolabeled antibodies were comparable. Although tumor uptake of the radioiodinated and radiometal labeled 3F8 was much higher than that of the radioiodinated F(ab')2 fragments (maximum percent injected dose per gram values were 39.4, 33.2 and 20.1 for 131I-3F8, 90Y-3F8 and 131I-F(ab')2, respectively), tumor to nontumor ratios were higher for radioiodinated fragments (with the exception of tumor to kidney ratio). For the minimum tumor dose necessary for complete ablation, the bone marrow received 195, 278 and 401 cGy for 131I-F(ab')2, 131I-3F8 and 90Y-3F8, respectively. Tumor doses were 50.1, 232 and 992 cGy/MBq for 131I-F(ab')2, 131I-3F8 and 90Y-3F8, respectively. Tumor to bone marrow dose, which is defined as the therapeutic index, was 21.5, 14.7 and 10.4 for 131I-F(ab')2, 131I-3F8 and 90Y-3F8. 131I-F(ab')2 fragments produced the highest therapeutic index but also the lowest tumor dose for radioimmunotherapy. Radiometal conjugated IgG produced the highest tumor dose but also the lowest therapeutic index.

Radioimmunotherapy of interleukin-2R alpha-expressing adult T-cell leukemia with Yttrium-90-labeled anti-Tac.

Adult T-cell leukemia (ATL) is a malignancy of mature lymphocytes caused by the retrovirus human T-cell lymphotropic virus-I. It is an aggressive leukemia with a median survival time of 9 months; no chemotherapy regimen appears successful in inducing long-term disease-free survival. The scientific basis of the present study is that ATL cells express high-affinity interleukin-2 receptors identified by the anti-Tac monoclonal antibody, whereas normal resting cells do not. To exploit this difference, we administered anti-Tac armed with Yttrium-90 (90Y) to 18 patients with ATL initially (first 9 patients) in a phase I dose-escalation trial and subsequently (second group of 9 patients) in a phase II trial involving a uniform 10-mCi dose of 90Y-labeled anti-Tac. Patients undergoing a remission were permitted to receive up to eight additional doses. At the 5- to 15-mCi doses used, 9 of 16 evaluable patients responded to 90Y anti-Tac with a partial (7 patients) or complete (2 patients) remission. The responses observed represent improved efficacy in terms of length of remission when compared with previous results with unmodified anti-Tac. Clinically meaningful (> or = grade 3) toxicity was largely limited to the hematopoietic system. In conclusion, radioimmunotherapy with 90Y anti-Tac directed toward the IL-2R expressed on ATL cells may provide a useful approach for treatment of this aggressive malignancy.

Radiolabeled anti-CD33 monoclonal antibody M195 for myeloid leukemias.

M195, a mouse monoclonal antibody reactive with the early myeloid antigen CD33, has been shown to target leukemia cells in patients and to reduce large leukemic burdens when labeled with 131I. A complementarity-determining region-grafted, humanized version (HuM195) has demonstrated similar targeting of leukemia cells without immunogenicity. We have studied two applications of therapy with 131I-M195. First, to intensify therapy prior to bone marrow transplantation (BMT), we combined 131I-M195 with busulfan and cyclophosphamide. Fifteen patients received first BMT for relapsed or refractory acute myelogenous leukemia or accelerated or blastic chronic myelogenous leukemia; four received second BMT for relapsed chronic or accelerated chronic myelogenous leukemia. Doses of 131I-M195 ranged from 120 to 230 mCi/m2. Few toxicities could be attributed to 131I-M195 therapy, and all patients engrafted. Eighteen patients achieved complete remission. Among those patients receiving first BMT, three have remained in unmaintained remission for 18+ to 29+ months. Six patients relapsed, including one with isolated central nervous system disease 32 months after BMT. Ten patients died in complete remission of transplant-related complications. Second, we studied whether 131I-M195 could reduce minimal residual disease and prolong remission and survival durations safely in patients with relapsed acute promyelocytic leukemia after they attained remission with all-trans-retinoic acid. Seven patients were treated with either 50 or 70 mCi/m2 131I-M195. Toxicity was limited to myelosuppression. As a measure of minimal residual disease, we monitored PML/RAR-alpha mRNA by reverse transcription PCR. Six patients had positive reverse transcription PCR assays prior to receiving 131I-M195; two converted transiently to negative. Median disease-free survival and overall survival of the seven patients were 8 (range, 3-14.5) months and 28 (range, 5.5-43+) months, respectively. This regimen compares favorably with others for relapsed acute promyelocytic leukemia. In an effort to avoid nonspecific cytotoxicity associated with 131I in future trials for minimal residual disease, we have conjugated short-range, alpha particle-emitting radioisotopes to HuM195 using a bifunctional chelate, 2-(p-isothiocyanatobenzyl)-cyclohexyldiethyl-enetriaminep entaacetic acid, with high efficiency and specific activities. 212Bi-HuM195 has demonstrated dose- and specific activity-dependent killing of HL60 cells in vitro. Injection of 213Bi-HuM195 into healthy BALB/c mice produced no effects on weight or viability.