Monoclonal antibodies (MoAbs) have the property of reactive specificity to tumor-associated antigens. In addition, their mammalian origin makes them nontoxic. They are, therefore, attractive candidates as therapeutic agents, either by themselves or conjugated to radionuclides, cytotoxic drugs, or toxins. The same specificity is exploited in radioimmunodetection, wherein the antibody is labeled with a suitable radionuclide that can be detected using current gamma camera imaging.
Radiolabeled antibodies have probably been studied in more clinical trials than any other form of immunoconjugate. The largest number of clinical trials with radiolabeled MoAbs, both diagnostic and therapeutic, have been carried out in patients with metastatic colorectal cancer . Fewer trials have been conducted in other solid tumors, notably, ovarian carcinoma , melanoma , lung carcinoma , renal cell carcinoma , breast cancer , carcinoma of the prostate , head and neck carcinoma , and central nervous system tumors. Immunotherapy with radioimmunoconjugates as well as immunotoxins has shown promise, especially in non-Hodgkin's lymphoma [9,10] and myelogenous leukemia . Phase I and II trials with radiolabeled antibodies in solid tumors have provided tantalizing responses, albeit none major [12-15].
All trials with murine antibodies, with the exception of those in B-cell lymphomas and pediatric neuroblastoma , have demonstrated that the probability of significant responses is hampered by the limitation of single use caused by their invariable immunogenicity. This has led to the development of nonimmunogenic forms, including chimeric  and humanized  antibodies, which will potentially permit multiple administrations.
Initial clinical trials of MoAbs focused on antibodies against carcinoembryonic antigen (CEA) [19,20], an antigen secreted by most differentiated adenocarcinomas. These trials, with iodine(Drug information on iodine)-131 as a radioconjugate permitting external detection (scintigraphy), revealed targeting of radioantibody to sites of known disease. They further demonstrated the safety of antibody administration. This led to the exploration of multiple "tumor-associated" antigens--secreted, cell-surface, and intracellular--in most tumor systems.
These trials have reached several conclusions:
- Monoclonal antibodies can be safely administered, with toxicity limited to rare allergic reactions.
- Toxicity of immunoconjugates is limited to the toxicity of the nuclide or toxin attached to the antibody. For radioimmunotherapy, this means that the critical organ is the bone marrow, with thrombocytopenia being the dose-limiting toxicity.
- Murine monoclonal antibodies usually evoke an immune response in humans, which precludes repeat administration, except in B-cell lymphoma, and in pediatric patients with neuroblastoma who have been heavily pretreated--both conditions in which immune compromise is considerable.
For radiolabeled antibodies, there are several additional considerations:
- Indium-111, used in radioimmunodetection, is limited in its ability to specifically target hepatic metastases owing to considerable nonspecific hepatic uptake of radioactivity, probably caused by trans-chelation.
- Technetium-99m, another radionuclide used for detection, will, owing to its short physical half-life of 6 hours, probably be more effective conjugated to antibody fragments, although rapidly targeting intact immunoglobulins are also likely to be useful.
- Radioimmunotherapy, especially for solid tumors, may require either large doses with marrow support or carefully planned fractionations.
- Beta-emitters, such as yttrium-90, iodine-131, and rhenium-186 are nuclides of choice for established disease; their major toxicity is hematopoietic. The low ionizing potential of these radionuclides probably limits their utility in adjuvant treatment, and their major role is probably in the treatment of residual small-volume disease.
Antibodies are members of a family of immunoglobulins produced in mammals by B-lymphocytes. The development of hybridoma technology by Köhler and Milstein in 1975 made possible the production of large quantities of MoAbs that recognize human tumor-associated antigens . Most currently used antibodies belong to the immunoglobulin G (IgG) subclass. The IgG molecule, with a molecular weight of between 150 and 180 kD, consists of two heavy and two light polypeptide chains connected by disulfide bonds. Pepsin splits part of the constant region (Fc) to produce an F(ab')2 fragment (~110-kD molecular weight). The enzyme papain splits the antibody molecule into one Fc fragment and two Fab' (~ 50 kD each) fragments. Figure 1 depicts the basic structure of the molecule, as well as the sites of action of pepsin and papain.
Radioimmunodetection studies have been carried out not only with intact immunoglobulin but with MoAb fragments as well. These have included F(ab')2 and, more frequently, Fab' fragments. The smaller the molecule, the faster its clearance from serum . However, Fab' fragments tend to be retained in the kidneys for extended periods, limiting their usefulness in the detection of disease around the kidneys. Moreover, fragments show less affinity for target antigens than intact immunoglobulin, and they are more difficult to produce and purify to clinical standard.
The radionuclide with energy emission characteristics most suitable for use with current nuclear medicine cameras is technetium-99m, which has gamma emissions of 140 KeV. This transitional element has no particulate emission; its short half-life further reduces radiation dose to the patient, permitting use of relatively large amounts of radioactivity. However, studies with technetium-99m-labeled MoAbs have been few . This is because fragment production and antibody (both immunoglobulin and Fab') labeling with technetium-99m are difficult procedures.
Rhenium is another transitional element with chemical characteristics comparable to those of technetium-99m. Rhenium-186 and -188 have been used in therapy. Both emit beta-minus radiation with concurrent gamma emissions, permitting imaging. Stability of binding of rhenium to antibody (usually carried out by an indirect linker) is an important consideration in the development of suitable rhenium-labeled antibodies in immunotherapy.
Chelation of MoAbs with radiometals, particularly indium-111, is difficult, with hepatic uptake of radioindium (presuMoAbly via trans-chelation in vivo to transferrin) limiting their utility in the detection of hepatic metastatic disease.24 The physical characteristics of indium-111 (half-life of 3 days, photons of 187 and 245 KeV, no particulate emission) make it an attractive nuclide for use in radioimmunodetection with intact immunoglobulin, despite its cost. Perhaps for this reason, the first radiolabeled MoAb to obtain FDA approval as a cancer-imaging agent was an indium-111-labeled MoAb, Oncoscint.
Yttrium-90, copper-67, and lutetium-177 are beta-emitting metals with therapeutic potential. Chelates do not bind radiometals to antibodies comparably. Lutetium-177 permits external imaging; yttrium-90 does not. Stability in vivo of therapeutic radiometals has not been optimal; advances in radiochemistry are encouraging, however.
Radioiodination of proteins, including MoAbs, is a relatively simple process, and thus, the greatest number of radioimmunodetection trials have been carried out with radioiodinated MoAbs. Iodine-123 has a short half-life (13 hours) and ideal emission characteristics (no particulate emission, 159-KeV photon emission). Its exorbitant cost has precluded its more widespread use, however, and its short half-life has deterred investigators from studying immunoglobulins labeled with iodine-123 .
Iodine-125 has low-energy photons that preclude external imaging but make it ideal for use in intraoperative radioimmunodetection (radioimmunoguided surgery) . Moreover, its long half-life of about 60 days permits its detection up to 6 weeks after administration, at which point normal tissue background levels are relatively low, permitting the detection of extremely small foci of abnormal antibody uptake. Consequently, most surgical radioimmunodetection studies have used iodine-125-labeled MoAbs.
Iodine-131 has a long half-life of 8 days and a complex decay scheme that includes beta-minus emission, precluding use of large amounts of radioactivity. This constraint, in addition to its high energy (364-KeV) gamma emission makes it less than optimal for gamma camera imaging, necessitating special collimation for current gamma cameras. Despite these limitations, however, the majority of radioimmunodetection trials have been carried out with iodine-131-labeled MoAbs . There are several reasons for this: (1) Iodine-131 is easily available and relatively inexpensive. (2) Protein radioiodination is relatively easy to carry out. (3) Radioiodinated MoAbs are relatively stable in vivo. (4) Persistence of free radioiodine in the body can be obviated by saturation of physiologic iodine stores with nonradioactive iodine (Lugol's solution or a saturated solution of potassium iodide(Drug information on potassium iodide)), permitting prompt clearance of free radioiodine by the kidneys without significant thyroid or stomach uptake; thus, nonspecific uptake of iodine-131-labeled MoAbs is not a problem. Moreover, the 8-day half-life of iodine-131 makes it possible to study the kinetics of antibody uptake by target tissue over time.
Successful clinical imaging of cancer was first reported using iodine-131 labeled to anti-CEA antibodies . It is particularly useful for the detection of neoplastic disease in the abdomen due to its low nonspecific uptake in the liver, spleen, and bone marrow. Iodine-131 has also been used in most radioimmunotherapy trials to date.
Regardless of the antibody used, dose-limiting toxicity has been myelosuppression. The high gamma emission entails hospitalization until radiation dose from the patient is sufficiently low (5 mR/h or less at 1 meter in the United States). Therapeutic trials with iodine-125 have been limited to MoAbs that internalize into the cell following interaction with antigen.