Use of radiolabeled monoclonal antibodies for diagnostic imaging

Monoclonal antibodies (MoAbs) are expected to carry radionuclides selectively to target tissues and to offer antigen-specific diagnosis. Indium (In)-111 has many favorable nuclear properties and is efficiently labeled with MoAbs using DTPA as a bifunctional chelating agent. In-111 labeled MoAbs are clinically employed for the diagnosis of malignant melanoma, colorectal cancer and acute myocardial infarction in Japan. Although non-specific deposit of In-111 was seen in liver and bone-marrow, scintigraphy using In-111 labeled MoAbs was encouraging, since it detected about 80% of tumors, tumors missed by conventional diagnostic methods such as CT, and tumors in patients with normal serum CEA values, and acute myocarditis as well as acute myocardial infarction was positive with In-11 labeled Fab fraction of anti-myosin Ab. Acute or subacute toxicity was not observed. Human anti-murine antibody (HAMA) was detected in 53 of 64 (82.8%) patients who were intravenously administered with 20 to 42 mg of anti-melanoma or anti-CEA MoAbs (whole IgG). In contrast, only 5 of 406 (1.2%) patients had detectable levels of HAMA in their serum after receiving 0.5 mg of Fab fraction of MoAb. Recently mouse-human chimeric Ab has been produced by recombinant DNA techniques, which localized well in xenografted tumors and seems to be promising for clinical use. Investigations are under way to increase the tumor to non-tumor ratio by modifying chelating agents for coupling MoAbs with radionuclides.     

Inhibition of autoradiolysis of radiolabeled monoclonal antibodies by cryopreservation

Autoradiolysis of therapeutic doses of monoclonal antibodies can occur rapidly, limits their shelf life and makes onsite radiolabeling a near-necessity. We evaluated freezing of three different 131I-labeled murine monoclonal antibodies at -70 degrees C, immediately following radiolabeling, as a method of diminishing autoradiolysis, and of preserving immunoreactivity. Freezing greatly limits the ability of radiation-induced free radicals to diffuse in solution and thus produce radiolytic damage. By freezing at -70 degrees C autoradiolytic damage of immunoreactivity of three different 131I monoclonal antibodies could be largely eliminated, in contrast to the 80-90% losses in immunoreactivity seen with storage at 4 degrees C for a period of 1 to 12 days. Reduced in vitro deiodination rates are also seen for frozen antibodies. Limited studies with 125I-labeled antibodies indicate autoradiolysis does occur, though at a slower rate per mCi than for 131I, and that this process is also retarded by freezing. Freezing may be valuable while quality control procedures are performed following radiolabeling as well as if temporary storage or shipment of radioantibodies prior to patient dosing is undertaken. While the approach should be validated for each antibody studied, freezing of therapeutic doses of monoclonal antibodies appears to be a simple and effective approach to the problem of autoradiolysis.     

Future role of radiolabeled monoclonal antibodies in oncological diagnosis and therapy

This review discusses the current limitations and future prospects of radiolabeled antibodies in cancer imaging (radioimmunodetection, or RAID) and therapy (radioimmunotherapy, or RAIT). Aspects such as the antibody vehicle, antigen target, radiolabel, tumor, host, and RAID and RAIT procedures are considered. In the short timespan for the development of RAID, tumors as small as 0.5 cm, which are sometimes missed by other radiological methods, can now be imaged with antibody fragments labeled with suitable radionuclides (eg, 111In, 123I, and 99mTc), particularly when single photon emmission computed tomography (SPECT) scanning methods are employed. 99mTc is clearly the preferred label, and the recent development of simple and rapid methods to attach this isotope to antibodies should be a welcome advance for the more widespread use of RAID. In RAIT, radiosensitive neoplasms, such as lymphomas, are already showing impressive responses to 131I-labeled antilymphoma murine monoclonal antibodies. Therefore, the successful conjugation of beta- and alpha-emitters to "humanized" monoclonal antibodies should provide a new generation of promising cancer therapeutics.     

In vivo kinetics of radiolabeled monoclonal anti-CEA antibodies in animal models

Studies were performed to determine the effect of the radiolabel and circulating carcinoembryonic antigen (CEA) on the pharmacodynamics of monoclonal anti-CEA antibodies (MoAbs). The studies were performed in normal BALB/c mice and in nude mice bearing human colon tumors. Three different tumors were used, each of which produced CEA levels characteristic of that particular tumor's secretory rate. The CEJ-326 MoAb labeled with either 111In or 125I was used in all studies. Circulating CEA induced the removal of 125I and 111In MoAbs from the vascular compartment. Liver concentrations of 111In increased and 125I levels decreased as the CEA secretory rate of the tumor rose. This indicates that circulating CEA complexes form in the vascular compartment which, in an animal model, are removed by the liver and spleen. This results in decreased tumor uptake of the labeled MoAb. The iodinated MoAb complexes are dehalogenated while the 111In is retained by the liver. This dehalogenation may account for the relatively low liver activity observed in radioimmunoimaging with intact radioiodinated anti-CEA MoAbs, provided the CEA complexes are similarly removed from the vascular compartment by the human liver.     

Pharmacokinetic study of radiolabeled anti-colorectal carcinoma monoclonal antibodies in tumor-bearing nude mice

Monoclonal antibodies (MoAbs) 17-1A and 19-9, which specifically bind human colorectal carcinoma (CRC) cells, were tested for their usefulness in localizing colorectal tumors in nude mice. One of the ¹³¹I-labeled MoAbs and an irrelevant ¹²⁵I-labeled immunoglobulin of the same isotype were injected into nude mice simultaneously bearing a human CRC and a human melanoma. The percentage of the injected dose of antibody per gram of tissue, the CRC/tissue ratios of antibody distribution, and the localization indices were calculated at various time intervals (2 h to 9 days). For both MoAbs, labeling to a specific activity of 10 Ci/g by the iodogen method gave optimum immunoreactivity. The accumulation of MoAb 17-1A in CRC reached its maximum at 5 days and remained at this level for up to 9 days postinjection. For MoAb 19-9, which detects a circulating antigen shed by the tumor into the serum, the accumulation in the CRC was maximum at 24 h, and decreased thereafter. The CRC/organ ratios and localization indices for both MoAbs increased with time in the CRC tissue, but remained low and unchanged in the melanoma and normal tissues. Using F(ab)₂ antibody fragments, faster kinetics with earlier maximum accumulation, higher tumor/organ ratios, and better localization indices were achieved than with intact MoAbs. The data obtained was useful in defining parameters which must be considered before radiolabeled MoAbs are used in cancer patients for diagnostic purposes.     

Distribution of radiolabeled human and mouse monoclonal IgM antibodies in murine models

The distribution and kinetics of six human and one murine monoclonal IgM antibodies (MoAb) were studied in BALB/c mice. Labeling was with 111In, 75Se, and 125I. The monomers and pentamers of certain MoAbs were studied. Human distribution studies were also performed. The serum containing [111In]MoAb was obtained from one of the patients 24 hr after administration and injected into mice which were then killed and assayed for 111In distribution. In general, the [75Se] and [111In]MoAbs had distribution and kinetic patterns that were similar while the 125I-labeled MoAbs dehalogenated after 4 hr. Monomers and pentamers had highly similar distributions suggesting that the distribution of IgMs may be based on factors other than molecular size. The murine IgM showed a somewhat different distribution in mice than did human IgMs. Serum from the patient containing [111In]MoAb had a distribution in mice similar to that of the patient with high liver and gastrointestinal uptake. The human imaging indicates that it is possible to target tumor with human IgM MoAbs, but significant problems remain in regard to their clinical use.     

Methods for the study of the metabolism of radiolabeled monoclonal antibodies by liver and tumor

Methods for elucidating the mechanisms by which radiolabeled antibodies are taken up and accumulated in tumor and liver are reviewed. These include the use of isolated perfused rat livers, RES blockade using dextran sulfate, single and double labeled antibodies, micropore chambers for the accumulation of the interstitial fluid, and in vitro tissue culture studies of antibody metabolism. Each method has its utility, examples of which will be discussed along with the methods' limitations. All of the methods have value in furthering our understanding of the metabolism of monoclonal antibodies both in vivo and in vitro. Use of these procedures to create a greater understanding of radiolabeled antibody metabolism, hopefully, will result in improved clinically useful agents for diagnosis and therapy.     

Preadsorption of radiolabeled monoclonal antibodies to liver and spleen tissues leads to higher tumor-to-normal-tissue ratios

This study addresses the impact of background activity on the use of radioimmunoconjugates for radioimmunodiagnosis and radioimmunotherapy. Since the liver and the spleen represent organs with preferential nonspecific uptake, we exposed radiolabeled (iodinated and Indium-111 labeled) preparations of monoclonal antibodies to a suspension of fresh liver and spleen cells at physiological temperature and compared their immunoreactivity, in vivo biodistribution, and tumor targeting to those of the same radiolabeled proteins without prior adsorption to this suspension. The biodistribution studies were performed under conditions of high background activity, i.e., shortly after the injection (1 hour) and using a high dose of the protein. Preadsorption of radiolabeled monoclonal antibodies results in a significant decreased uptake in certain normal tissues, i.e., greater contrast between normal and tumor tissues, as demonstrated by the quotient of the two target-to-nontarget ratios (exposed/unexposed antibody) which was greater than one for most of the tissues examined.     

Immunoscintigraphy of human tumors transplanted in nude mice with radiolabeled anti-ras p21 monoclonal antibodies

Anti-ras p21 monoclonal antibody (RASK-3) was used for immunoscintigraphy of human cancer cell lines in nude mice. Iodine-125-labeled RASK-3 was injected into nude mice with either human colon cancers (FCC-1 or BM-314) or lung cancer (KNS-62). Clear images were obtained in all three cancers 7 days after the injection of antibody. No localization of 125I-labeled control monoclonal antibody was observed. The ratio of tissue/blood radioactivity and % ID/g in the tumor were significantly higher than other organs by Day 8. The specific localization index examined by 131I-RASK-3 and 125I-control monoclonal antibody was also higher in the tumor than in other tissues. In the in vitro study, binding of RASK-3 to tumor cells increased significantly by treatment of cells with either lysolecithin or periodate-lysine-paraformaldehyde, which confirmed the intracellular localization of ras p21. The mechanism by which anti-ras p21 antibodies accumulate in tumor sites could be the necrotic changes in tumor cells or changes in membrane permeability of non-necrotic cells. These results provide a strong rationale for the utilization of ras p21 as a target antigen in the imaging of a variety of human cancers.     

Relationship between in vitro binding activity and in vivo tumor accumulation of radiolabeled monoclonal antibodies

The relationship between in vitro cell binding and in vivo tumor accumulation of radiolabeled antibodies was studied using 125I- and 111In-labeled monoclonal antibodies to human osteosarcoma, and a human osteosarcoma xenograft (KT005) in nude mice. Three monoclonal antibodies--OST6, OST7, and OST15--raised against human osteosarcoma recognize the same antigen molecule. Although the binding of both 125I- and 111In-labeled OST6 to KT005 cells was higher than that of radiolabeled OST7 in vitro, 125I-labeled OST6 showed a faster clearance from the circulation and a lower accumulation in the transplanted tumor than 125I-labeled OST7. In contrast to the radioiodinated antibodies, the in vivo tumor accumulation of 111In-labeled OST6 was higher, although not significantly, than that of 111In-labeled OST7. OST15 showed the lowest binding in vitro, and its in vivo tumor localization was also lower than the others. The discrepancy in tumor uptake between OST6 and OST7 labeled with either 125I or 111In may have been a result of differing blood clearance. These results suggest that binding studies can be used to exclude from in vivo use those antibodies which show very poor binding in vitro, while in vivo serum clearance may be a better test for choosing antibodies with similar binding.     

Altering pharmacokinetics of radiolabeled antibodies by the interposition of metabolizable linkages. Metabolizable linkers and pharmacokinetics of monoclonal antibodies.

Monoclonal antibodies, their fragments and low molecular weight oncophilic molecules such as synthetic somatostatin derivatives have been used to deliver radioactivity to target cells for both diagnostic and therapeutic purposes. Clinical studies demonstrated high abilities of the radiolabeled antibodies and peptides for nuclear medicine applications. However, high and persistent localization of radioactivity was observed in the liver or kidney especially when these molecules are labeled with metallic radionuclides, which reduce diagnostic accuracy and compromise therapeutic effectiveness. Thus, radiolabeled antibodies and peptides would become much more useful in both targeted imaging and radiotherapy if the undesirable radioactivity localization can be diminished. As a means to reduce the undesirable radioactivity, interposition of a metabolizable linkage between an antibody and a radiolabel was proposed to generate radiolabeled small molecules of urinary excretion from the parental antibody by enzymatic cleavage of the linkage. In this paper, after indicating the rationale behind the radiopharmaceutical design, a significant role played by the interposition of the metabolizable linkage in altering pharmacokinetics of radiolabeled antibodies is described from a variety of studies so far reported with an emphasis being laid on the importance of radiometabolite-based design of metabolizable linkages.     

Imaging of colorectal carcinoma with radiolabeled antibodies

Colorectal cancer has been the tumor type most frequently studied with radiolabeled antibodies. Among the various antibodies, a majority of patients with colorectal cancer have received xenogeneic polyclonal or monoclonal antibodies against carcino-embryonic antigen. This review summarizes the current status of colorectal cancer imaging with radiolabeled antibodies, ie, radioimmunodetection (RAID), and examines the published studies involving carcinoembryonic antigen (CEA) antibodies and 17-1A, 19-9, and B72.3, and other monoclonal antibodies. In order to better address the issue of the current and future clinical usefulness of this emerging technology, particular attention is given to the protocols, methods, and results of the published studies. Despite differences in study parameters, antibodies and forms, labels, administration routes and doses, and scanning instruments and methods, it has been found that (1) almost no adverse reactions have been evident; (2) antibody fragments are preferred over whole immunoglobulin G reagents because they achieve higher tumor-to-background ratios earlier, thus reducing or precluding the need for dual-isotope subtraction methods or long delays before imaging; (3) use of antibody fragments, including the monovalent Fab' form, permits imaging with short-lived radionuclides of excellent photon properties, such as 123I and 99mTc; (4) circulating antigens against which the imaging antibody is directed can complex with the injected antibody, but such complexes have not prevented successful RAID; (5) patients with high serum titers of the appropriate antigen target usually have higher rates of positive RAID; (6) patients who are seronegative for the tumor antigen being studied can have positive RAID findings, which can represent the detection of occult lesions; (7) single photon emission computed tomography appears to provide better image resolution than planar scanning; (8) regardless of the sensitivity reported in any particular study, almost all investigators have observed the disclosure of occult neoplasms by RAID; and (9) RAID, a more functional test of usually high specificity, can complement other radiological methods, such as computed tomography scans, which are limited to structural information.