Principles of Molecular Targeting for Radionuclide Therapy
Molecular targeting requires assessing several factors that come into play such as the location of the target, the choice of radionuclide, the inertness of the bifunctional chelate and stability of the covalently bound halogens, matching the residence time in the tumor with the physical half-life of the radionuclide, the scale and scope of the disease, and the absorbed dose sensitivity of the targeted tumor compared to normal tissue. The principles of molecular targeting are well established, but a paradigm shift from designing a medium-affinity radiotracer used to determine target density to designing a high-affinity, high-target density radioligand to maximize the target-to-nontarget ratio should increase the probability of detecting lesions smaller than the instrument resolution.
Developing and validating a therapeutic radiopharmaceutical for a single target is necessary, but often not sufficient to produce a toxic event because of other mechanisms that are only partially understood. These include nontargeted effects due to radiation emitted from neighboring, targeted cells as well as bystander effects produced by the cellular processing of radiation not necessarily impinging on DNA. Both of these indirect consequences of cellular radiation could make a substantial contribution to the efficacy of targeted radionuclide therapy. These mechanisms should be exploited to optimize the efficacy of targeted radiotherapy and overcome the inefficiency of tumor control due to nonuniform distribution of radiation dose. The design approach to take advantage of the indirect consequences of cellular radiation depends heavily on further elucidation of the indirect effect. The successful combination of these two should lead to more effective nuclear radiotherapy.
KeywordsAlpha particles Beta particles Auger electrons Molecular targeting Therapeutic radionuclides Theranostics Double-strand breaks Bystander effects Residence time
Ataxia telangiectasia mutated
ATM and RAD3 related
Area under the time-activity curve
Cerebral blood flow
Cross-bridged macrocyclic chelators
X-ray computed tomography
European Medicines Agency
Enhanced permeability and retention
Fluorescence-activated cell sorting
United States Food and Drug Administration
Green fluorescent protein
Gray unit (ionizing radiation dose in the International System of Units, corresponding to the absorption of one joule of radiation energy per kilogram of matter)
Cell line of human colon carcinoma
Human epidermal growth factor receptor
Linear energy transfer
Lymph node metastasis from carcinoma of the prostate
Colon adenocarcinoma cells line name
Muscarinic acetylcholine receptor
Myocardial blood flow
Michigan Cancer Foundation-7, a breast cancer cell line
Molecular Insight Pharmaceuticals
Messenger ribonucleic acid
S-2-(4-Isothiocyanatobenzyl)-1,4,7,10-tetraazacyclododecane tetraacetic acid
Cross-bridged (propyl) TE2A
Positron emission tomography
Prostate-specific membrane antigen
A 5′ to 3′ DNA helicase involved in nucleotide excision repair and transcription (from “RADiation sensitive”)
Radiation-induced biological bystander effects
Reactive oxygen species
Single-photon emission computed tomography
Single-photon emission computed tomography/computed tomography
Standardized uptake value
Transfectant mosaic spheroids
Transfectant mosaic xenograft
Therapy operating characteristic
Technetium, 2-[[2-[[[3-(4-chlorophenyl)-8-methyl-8-azabicyclo[3.2.1]oct-2-yl]methyl](2-mercaptoethyl) amino]ethyl]amino]ethanethiolato(3-)-oxo-[1R-(exo-exo)]
Zirconium desferrioxamine B
Potential targets for radioligands
Receptor is defined as a molecule in a cell membrane that responds to a particular neurotransmitter, hormone, antigen, or other substance. To facilitate imaging, radiolabeled antagonists to the receptor rather than lower-affinity, natural ligands are usually employed
Acceptor is defined as a molecule in a cell membrane whose biological activity is not known. The mannose receptor in the liver and the prostate-specific membrane antigen are examples
Enzyme is defined as a substance that acts as a catalyst to bring about a specific biochemical reaction. Inhibitors of the enzyme as often used, for example, radiolabeled fludeoxyglucose at hexokinase, but substrates as well, for example, 18F-DOPA is used to quantify DOPA decarboxylase activity in the brain
Transporter protein is defined as a protein that serves the function of transporting ligands within the cell
Protein antigen is a specific antigen, usually cell surface, that is unique to the tumor. Antibodies and antibody fragments are produced by recombinant and proteolytic methods
Potential platforms used in radioligand design
Small molecules are chemical entities that have a molecular weight of <500 Da. They are especially used in neurochemical applications given the requirement to cross the blood-brain barrier
Nanoparticles are usually defined as having a diameter of 100 nanometers or less. They are used in situations where the number of ligands for either diagnostic or therapeutic use can be bound to the surface
Antibodies and fragments: scFv (~28 kDa), diabody (~55 kDa), minibody (~80 kDa), F(ab’)2 (100 kDa), and IgG (~150 kDa). Various proteolytic or recombinant methods result in a combination of the binding region that targets the tumor cell antigen and other fragments present in the IgG
Enhanced permeability and retention (EPR)
This does not involve a single target, but a difference in EPR in target and normal tissue. In the 1980s, Maeda and colleagues found that macromolecules such as polymers and proteins with molecular weights larger than 40–50 kDa showed selective accumulation in tumor tissues, far more than that observed in normal tissues; moreover, they retained in tumor tissues for long periods, i.e., >24 h. They coined this unique phenomenon enhanced permeability and retention (EPR) effect. Accordingly, an EPR-based tumor-targeting strategy (macromolecular therapy) was developed by using polymer modification, nanoparticles, micelles, liposome, etc.
Except for radionuclides of iodine, bromine, and astatine, i.e., halogens , therapeutic radionuclides are metals and require a bifunctional chelate conjugated to the targeting molecule. Halogen radionuclides most often are linked via a residualizing group . Based on the cellular location and density of the target, the design of the targeted radiotracer will follow the approach for nuclear diagnostics with certain changes required for therapeutic radiopharmaceuticals [7, 8]. Relevant properties for a diagnostic radiopharmaceutical are an inexpensive radionuclide, high radiochemical yield, high radiochemical purity, and high specific activity including effective specific activity. These are required for therapeutic radiopharmaceuticals as well, although a high specific concentration of a therapeutic dose will produce radiolysis unless a radical scavenger is included. High specific activity is still a necessity to avoid saturation of the target and pharmacologic effects given the higher mass dose. For diagnostics, given that external imaging detects photons and not chemical structure, important criteria in vivo are metabolic stability; efficient transfer to the target only, i.e., target specificity; and pharmacokinetics that are more heavily weighted by biochemistry than by delivery although there is an ongoing paradigm shift to favor maximal delivery combined with optimal specific uptake and retention. The same criteria apply for therapeutic radiopharmaceuticals although pharmacokinetics heavily weighted for efficient extraction fraction may have an advantage in delivering a toxic dose to a larger percentage of cancer cells. A higher input area under the curve is also important for therapeutic radiopharmaceuticals. For positron-emitting radionuclides, the distinction between what constitutes diagnostic and therapeutic injected radioactivity is often blurred by the focus on the external emissions of positron annihilation rather than the high LET of the emitted positron.
The major challenge is to deliver the therapeutic radiopharmaceutical to the target whether it is on the cell surface, in the cytosol, or in the nucleus or more generally whether it is a well-characterized target or a phenotypic target obtained by screening . The latter is most often the approach for antibodies obtained by screening antibodies and fragments to the tumor cell surface for specificity and high affinity. Another challenge is to demonstrate binding to the chosen target in the absence of specific and nonspecific binding to other sites using an array of drugs known to bind to the specific target or to closely related targets including subtypes. Using target-specific knockout mice can also demonstrate specificity [10, 11]. In general, the probability of achieving the therapeutic goal decreases when moving from targets on the cell surface to targets in intracellular compartments.
Another major challenge in radiopharmaceutical design requires a clear approach to the optimal pharmacodynamics, i.e., the ideal combination of flow and permeability, which constitute delivery and kinetics at the target, which is based on the on-rate and the off-rate of radiopharmaceutical binding. The early diagnostic radiopharmaceuticals had very high affinities leading to high target retention to maximize the potential for external imaging, but with this came pharmacokinetics that are heavily delivery related. For example, diagnostic radiotracers such as N-[11C]methylspiperone  and 3-quinuclidinyl-4-[123I]iodobenzilate  (IQNB) have high affinities and high retention, and therefore their biodistribution is heavily delivery related. For IQNB, given the longer half-life of 123I, the distribution in vivo with time shifted to dependence on muscarinic receptor density . Investigators most often used IQNB to determine if new drugs involve the muscarinic receptor . Experimentally, the ideals of delivery-related radiopharmaceuticals are those that are trapped in capillaries smaller than the particle size. For example, microspheres administered intravascularly, by virtue of efficient trapping in the capillaries, are the gold standard for heavily flow-related agents. In the clinic 99mTc-labeled microspheres are used to measure lung perfusion in vivo. The therapeutic application that best matches microspheres is intra-arterial radiotherapy, using radiolabeled 90Y microspheres of glass or resin to treat lesions in the liver . Small molecules can approach the extraction efficiency of microspheres if the extraction and retention of the injected radiopharmaceutical in the tumor is >90%, but given the cardiac output to the target after IV injection, only the lung will retain a high percentage of the injected dose.
Most of the diagnostic radiotracers developed subsequently were “more reversible” with slightly lower affinity so although the original distribution might be weighted toward delivery, subsequently (within the time frame of the radionuclide half-life), the distribution would be proportional to the biochemistry involved. Separating flow, permeability, and target binding has been a challenge because rapid and efficient uptake is advantageous for both diagnostic and therapeutic radiopharmaceuticals from a total delivery point of view; therefore, the measure of changes in biochemistry as a function of disease often required a measurable off-rate from the target.
But for diagnostic radiotracers, it is clear that efflux of radioligand not specifically bound to the target in the tissue compartment must be rapid and efficient within the time frame of the radionuclide half-life. In cardiology, radioligands designed to measure changes in biochemical pathways are heavily weighted by distribution and require saturation studies in vivo to determine adrenoceptor density . One example of a reversible ligand from neurochemistry studies is 3-[4-[(3-[18F]fluoropropyl)thio]-1,2,5-thiadiazol-3-yl]-1,2,5,6-tetrahydro-1-methylpyridine (18F-FP-TZTP), whose early distribution correlated with perfusion as measured by [15O]H2O, i.e., within the range of cerebral blood flow (CBF) values measured, there is excellent agreement between CBF and the K1 of 18F-FP-TZTP . But the later distribution was weighted by a combination of muscarinic receptor density and synaptic acetylcholine concentration [19, 20]. The goal of the “reversible” ligands is to determine the change in target density as a function of disease or treatment. But is this the most valuable use for nuclear diagnostics with the inherent goal of whole-body imaging?
Ideal properties of diagnostic and therapeutic targeted radioligands
Bmax fmol/mg protein (nM)
SERT (frnt cortex)
NET (frnt cortex)
Although the location of the target (extra- versus intracellular) is a major design criterion, the development of a therapeutic radiopharmaceutical also depends on the clinical need and the subsequent location of the key target and on a number of other design criteria, which are discussed in the following sections.
The Choice of Radionuclide
Choosing the radionuclide is dependent not only on the location of the target (cell surface, cytosol, nucleus, specifically RNA and DNA) but also on matching the radionuclide physical half-life with the biological half-life of retention at the target. For example, the choice of half-lives for directly labeling antibodies binding to a cell surface target is of the order of days not hours. Matching the emission properties to the scale and scope of the disease being treated is also of critical importance.
Radionuclides of current interest in targeted therapy
The approved radionuclide therapies by the FDA and the EMA include iodine-131 iodide for differentiated thyroid cancer, strontium-89 chloride (Metastron) for bone pain, samarium-153 lexidronam (Quadramet) for bone pain, yttrium-90 ibritumomab tiuxetan (Zevalin®) for NHL, iodine-131 tositumomab (Bexxar) for NHL, radium-223 dichloride (Xofigo) for bone metastases, and 177Lu-DOTA-octreotate (Lutathera) for NE tumors. It is interesting that [131I]iodide and [223Ra]radium dichloride are most often used in the clinic. It may be that the unmet need of these two radiotherapeutics at the time of FDA approval was a key metric for their clinical impact.
Generators, as has been the case for the 99Mo/99mTc generator (67-h half-life of 99Mo and 6-h half-life of 99mTc), would likely be the most economical if the pharmaceutical requirements, namely, retaining apyrogenicity, sterility, radionuclide yield and purity, and effective specific activity, can be met over the lifetime of the generator. 68Ge/68Ga (271-day half-life of 68Ge and 68-min half-life of 68Ga) is the generator most often used in PET diagnostic procedures and has been registered in Europe. Generators for therapeutic radiopharmaceuticals will most likely be used by central radiopharmacies given the pharmaceutical requirements, but the half-lives of the parent nuclide, e.g., 225Ac (with 10-day half-life) for 213Bi (with 46-min half-life) and 212Pb (with 10.6-h half-life) decaying to 212Bi (with 60-min half-life), will limit shipping range.
Bismuth-213 decays by beta minus emissions (97.9%) to the very short-lived 213Po and 2% by alpha decay to 209Tl. This is interesting because if the 213Bi is purified and chelated, the daughter has such an extremely short half-life that it will remain in the vicinity of the 213Bi .
Astatine-211 has been used most often alpha emitter. The challenge is the chemistry of astatine because there is no stable isotope to carry out the necessary analytical studies. Astatine-211 has a half-life of 7.2 h. It decays with 58.3% probability through the electron capture process to 211Po, which has a half-life of 0.52 s, and decays by emitting a 7.45 MeV alpha particle to stable 207Pb. As a consequence of the electron capture process, polonium K X-rays are emitted that permit external imaging, and the two most abundant of these X-rays have energies of 77 and 80 keV with yields of 0.1 and 0.2 per disintegration .
Which of the therapeutic radionuclides have gamma-emitting radioisotopes that could be used in a theranostic application? In this situation, it is important to have the same metal (or halogen) for both applications to maintain identical chemistry. For example, it appears that 111In as a gamma-emitting partner for 90Y is not acceptable, whereas 86Y is . 211At does not have a high-abundance gamma-emitting radioisotope (nor a stable isotope either), and so surrogates are necessary. The fission product metals generally do not have gamma-emitting radionuclides for imaging although some daughters emit gamma rays. But in decay schemes with multiple radioactive daughters, it is best if the gamma rays are emitted with the decay of the parent and not the decay of the daughters. Suggestions have been made to use bremsstrahlung imaging for 90Y, which only emits high-energy annihilation gamma rays at very low abundance [38, 39].
Instrument Resolution and Biological Resolution
For theranostics, whether the use is in pre-approval studies to choose the optimal therapeutic and to monitor the success of that therapy, the goal of imaging smaller-volume lesions using quantitative emission tomography is crucial. External nuclear imaging is unique in its ability either to identify the target on cancer cells on a whole-body basis or to monitor the number of tumor cells using a radiolabeled biomarker that is targeting a high-density site that is independent of the treatment paradigm. The challenge of monitoring the same site targeted by the therapeutic is in the interpretation of a decrease in targeted radioactivity because a decrease could represent changes in the number of targets per tumor cells or a decrease in the number of tumor cells with the number of targets per cell constant or blocking by the mass dose of the therapeutic. This latter is less of a challenge in high specific activity therapeutic radiopharmaceuticals, but even in that case, achieving the maximal tumor dose may very well represent enough mass to occupy more than 5% of the target.
Monitoring the extent of the disease to include small lesions that are smaller than two to three times the instrument resolution requires a high target-to-nontarget (T/NT) ratio. Counting statistics, spatial resolution blurring, and target-to-background ratios influence the detection of small volumes by external imaging. In one example from 1999, Raylman et al. determined the detectability of small objects using a PET camera of that era with a resolution of ~9 mm . In phantom studies, they found that spheres were detectable when larger than 9 mm in diameter if the target-to-background ratios exceeded 3:1. For smaller-diameter spheres, larger target-to-background ratio was required (18:1) to overcome the resolution blurring and the statistical uncertainty of background radioactivity. However, endogenous ligands radiolabeled with 11C and radiofluorinated analogs of glucose, thymidine, choline, and acetate, along with other radioligands such as 18F-fluoromisonidazole, 64Cu-ATSM, and 18F-florbetapir, have standardized uptake values (SUV) of 3 or lower for the cutoff between normal and abnormal lesions.
The lower resolution of SPECT can be countered with higher target-to-nontarget ratios for diagnostic radiopharmaceuticals. Furthermore, the higher target-to-nontarget ratios favor efficacious monitoring of therapeutic radiopharmaceuticals as described above. Togawa et al. imaged phantoms without scatter and attenuation corrections using a Toshiba three-head rotating SPECT system filled with 99mTcO4 −. The FWHM of the SPECT system in this study was 13 mm, and the target-to-nontarget ratio obtained from the image of the phantom with a diameter of 46 mm (3.5 times that of the FWHM) shows a count recovery of 96.7 ± 6.2% of the actual target-to-background (T/B) ratio in the phantom as measured in a gamma counter. In contrast, the count recovery was 81.3 ± 7.2% and 68.1 ± 5.5%, respectively, even when the diameter was 37 mm (2.8 times that of the FWHM) and 29 mm (2.2 times that of the FWHM). For actual target-to-background ratios in the phantom ranged from 2.0 to 8.1 as measured in a gamma counter, the correlation coefficients increased from 0.037 to 0.165, respectively. The authors also reported the target-to-background ratio (T/B) multiplied by the diameter (d) to correct for the partial volume effect. Also, there was a significant linear correlation between T/NT • d obtained from the images and the actual T/B obtained from the gamma counter .
Recently, Zeintl et al. demonstrated the ability to obtain accurate data from phantoms in vitro and in patients after injection of 99mTc diphosphonate using SPECT/CT. Although the accuracy in phantom studies and measurements in vivo of radioactivity in the bladder were within 5%, the standard error reflects a need to increase the precision even further [42, 43].
A recent publication using a 68Ga radioligand for targeting PSMA gave tumor-to-gluteal musculature ratios at 3 h of 28.3 with a wide range between 2.9 and 224.0. These were based on 65 lesions considered typical for metastases or local relapse. The metastases were present in the bone and lymph nodes. Factors such as the radioactivity injected, the mass injected, the PSA value, the size of the lesion, partial volume losses, etc. were not correlated with the ratio by the authors. In a patient study, the 68Ga ligand uptake was detected in more tumors than were detected with 18F-fluoroethylcholine, which has a lower tumor-to-background ratio .
Radiobiological Properties Affecting the Choice of Radionuclide
The most significant limitation of targeted radionuclide therapy is the heterogeneity of binding of radiopharmaceuticals in tumor deposits. Attempts have been made to overcome the nonuniformity of radiopharmaceutical uptake that results in underdosing of some tumor regions by the application of radionuclides that decay to emit long-range particles, so that neighboring nontargeted cells might receive a radiation dose by cross fire. Even with long-range radionuclides such as 131I or 177Lu, because their emissions are of low linear energy transfer (LET), subpopulations of tumor cells might experience less than a sterilizing dose.
Radiation Microdosimetry and Particle Range
Decay particles have specific energy spectra and capacities for tissue penetration. Some features of therapeutic radionuclides are presented in Table 4. The finite ranges of emitted particles give rise to radiation microdosimetry considerations which are unique to targeted radiotherapy.
Most clinical experience has been acquired with beta emitters whose mean particle range is equivalent to several cell diameters. Therefore, the radiation dose to individual cells is the sum of cross fire – radiation emanating from neighboring cells which have accumulated radiolabeled agents – and radiation from radionuclide concentrated within the cell itself. In the case of radionuclides that decay to emit beta particles , cross fire is the most significant constituent of dose received by any cell. Conversely, no cross fire is associated with the uptake by cancer cells of Auger electron emitters. Accordingly, high-LET Auger therapy, while expected to be tumor specific, is anticipated to provide no compensation for heterogeneity of targeting. Then again, emerging evidence suggests that radioactive emission from targeted cells is not the only bystander effect operating in radionuclide treatment of cancer. It is becoming clear that radiation-induced biological bystander effects (RIBBEs) deriving from the cellular processing of the physical radiation insult, which need not interact directly with DNA, may play an important part in the overall efficacy of radionuclide targeting (see section “Radiation-Induced Biological Bystander Effects (RIBBEs)”).
The dimensions of the tumor itself are very important. For micrometastases whose diameter is less than the range of decay particles, absorption of radiation energy is inefficient. Indeed a significant fraction of this energy is deposited outside the tumor . Hence, microtumors will be underdosed and more resistant to sterilization. As tumor size increases, more decay energy is absorbed and the probability of cure is enhanced. However, as tumor size increases beyond particle range, the absorption efficiency reaches a plateau. Furthermore, the presence of a larger number of cells decreases the likelihood of successful treatment.
Such considerations were responsible for the concept of an optimal tumor size for maximal curability by specific radionuclides. Mathematical modeling has demonstrated that the optimal tumor diameter is slightly greater than the mean particle range of the radionuclide. For example, in the case of 131I, homogeneously distributed throughout a tumor, the ideal diameter range for curability is 2.6–5.0 mm [46, 47]. However, with increasing tumor size, the uptake of radionuclide will be increasingly nonuniform, resulting in greater resistance to radionuclide therapy of macroscopic tumors.
These predictions are supported by experimental evidence of the differential susceptibility to growth inhibition of multicellular tumor spheroids of a range of sizes treated with metaiodobenzylguanidine (MIBG) labeled with [123I]iodine, [125I]iodine, [131I]iodine, or [211At]astatine . MIBG is selectively accumulated by an active mechanism in cells of neural crest origin [49, 50], and conjugates of 123I and 131I are used for the imaging and therapy, respectively, of neuroendocrine tumors.
The Auger electron-emitting conjugates ([123I]MIBG and [125I]MIBG) and the alpha-emitting conjugate ([211At]MABG) were highly toxic to small spheroids, whereas the beta-emitting conjugate [131I]MIBG was relatively ineffective. Furthermore, the Auger electron emitters were more effective than expected, given the extranuclear localization of MIBG. As dosimetrically predicted, however, [211At]MABG was extremely potent in terms of both concentration of radioactivity and number of atoms per unit volume administered. In contrast, the Auger electron emitters were ineffective in the treatment of larger spheroids, while the beta emitter showed greater efficacy . These findings suggest that short-range emitters would be well suited to the treatment of circulating tumor cells or small clumps, whereas beta emitters would be superior in the treatment of subclinical metastases or macroscopic tumors. Such experimental results provide support for a clinical strategy of combinations (“cocktails”) of radioconjugates in targeted radiotherapy.
The Choice of a Chemical Platform
Salts and Covalent Bonds
Some of the diagnostic and therapeutic radionuclides are simple salts: sodium ([123/131I]iodide), [67/68Ga]Ga citrate, 201Tl chloride, 223Ra dichloride, and Na [18F]fluoride. Others are covalently bound to the targeting molecule, for example, meta[123/131I]iodobenzylguanidine (MIBG). Likewise, 211At is covalently bound to the targeting molecule. In all cases the stability of the covalent bond is crucial given that in diagnostic agents, the emitted photons do not reveal the chemical structure of the radiotracer and in therapeutic applications it is important that the radionuclide remains with the targeting molecule for maximal therapeutic index .
To develop a true tracer of a targeting peptide or protein using a bifunctional chelate, the metal must be kinetically inert to hydrolysis and exchange labeling with endogenous proteins in vivo. In addition, metal targets to produce the radioactive metal, solvents, and all containers are potential sources of nonradioactive metals, which can reduce the effective specific activity and lead to possible saturation of a targeted site . The same is true for the residualizing agents labeled with radiohalogens. The effective specific activity of [76Br]bromide can also be lower than the specific activity itself given impurities in the metal target as well as the reagents used in the purification .
Brechbiel has pointed out that many chelating agents are misnamed because one of the functional groups has been used to link to the targeted molecule. His recommendation is to link to the targeted molecule using a carbon atom in the backbone to maintain the full functionality of the chelating moiety. He has demonstrated the relative inertness for both acyclic and cyclic (macrocyclic) chelating agents using the latter approach .
Endogenous proteins are strong competitors for Ga, Zr, and Cu. Transferrin has a high affinity for Ga. The Ga in Ga citrate is thought to be transferred to transferrin after injection, and therefore Ga transferrin is the targeting entity. There may be other proteins that bind Ga as well such as lactoferrin and serum proteins .
89Zr-desferrioxamine B (DFO) appears to be stable to metal exchange with serum in vitro, but in vivo Zr-DFO shows bone uptake, but significantly less than Zr-citrate . In vivo, transferrin may also be involved in trans-chelation of Zr. Although transferrin receptor upregulation has been demonstrated in specific tumors, enhanced permeability and retention (EPR) effect  could also be a factor in the uptake . Earlier studies comparing 18F-albumin and 18F-transferrin in a tumor model containing upregulated transferrin receptor reported higher uptake with albumin rather than transferrin illustrating that the EPR effect could produce higher uptake compared to receptor binding. With the 18F label, the transferrin and albumin were directly assayed by radiolabeling the protein rather than via the chelated metal .
The copper transporter has drawn interest because Cu salts appear to localize in cells via the copper transporter. If the Cu can be transferred to the Cu transporter present in hepatocytes and upregulated in tumor cells, this constitutes another pathway to use Cu radionuclides in diagnostic or therapeutic radiopharmaceuticals. The earliest report appears to be from Apelgot et al., who found that 42% of the 64Cu in mouse ascites fluid after i.p. injection of [64Cu]CuCl2 was isolated with the DNA fraction. The electron capture emissions are effective if the metal is associated with DNA. Then the nuclear transformation (transmutation) with the associated changes with decay can produce a lethal event . Others have followed these early studies with in vivo studies in tumor-bearing mice [63, 64, 65, 66].
There is an interesting analogy with the chemotherapeutic, cis-diamminedi-chloroplatinum(II) (cis-Pt), which enters cells via the copper transporter and then binds to the guanine bases in DNA after the two chlorine atoms hydrolyze in the lower pH milieu. Cross-linking DNA through either inter- or intrastrand cross-links seems to be the mechanism of action. Wolf carried out experiments with 195mPt cis-platinum and reported that the 195mPt catabolite was bound irreversibly to plasma proteins . The subcellular distribution %ID/g (x10−2) of 195mPt 4 h after injection of 195mPt cis-platinum in tumor-bearing rats was 0.43 in DNA, 4.13 in chromosomal proteins, and 4.2 in cytosolic ligands . However, the DNA binding seems key to the therapeutic action of the drug.
Whether a copper salt will be as effective as a bifunctional chelate containing radiocopper is a question not yet answered. Certainly, the copper salt is taken up in the liver via the receptors in hepatocytes, whereas the biodistribution between the liver and kidney for the bifunctional chelates can be altered, for example, by changes in the linker between the chelate and the targeting molecule.
There is some question of the therapeutic index of fission product radionuclides if through multiple decays the daughter radionuclides do not remain with the chelating agent. Boswell and Brechbiel report a general percentage of released radionuclide to be 30% . The loss of decay product daughters from beta decays is variable. Recently a publication on 44mSc/44Sc indicates that there is no release during decay . And, of course, catabolism of the linker between the chelating agent and the targeting molecule must be stable to various enzymatic and hydrolytic processes.
Another key consideration for bifunctional chelates is whether the perturbation caused by the linker and metal chelate alters or hinders binding to the target. The tracer principle at least for the biochemistry required for the diagnosis or therapy must be validated. Concern for the perturbation caused by the metal chelate is most often reported, but simple replacement of a hydrogen or hydroxy group with fluorine is not without consequences [70, 71].
In summary, since external imaging for diagnostics and therapeutic index are based on localization of the injected radiopharmaceutical, a kinetically inert covalent bond for the halogens and an inert metal chelate are crucial. It is key for therapeutic radiopharmaceuticals that the radionuclide remains as part of the targeted bifunctional chelate to maximize tumor damage and minimize normal tissue damage. The bystander effect is important. In small animals it could represent a significant effect that won’t be characteristic of the human distribution. Furthermore, the lesser effect in Auger electron therapy may necessitate the need for nuclear uptake in a higher percentage of cells.
The Location of the Target: The Challenges of Designing Radiotracers for Intracellular Compartments
Cell surface targets represent a large percentage of the therapeutic targets to date. Internalization of the cell surface target is an advantage if it is a residualizing radiopharmaceutical, that is, it remains within the cell. Metal chelates have been superior to iodinated compounds in that respect because if iodide or iodotyrosine is a catabolic product, the efflux from the cell will be significant. For metal chelates, anion production is less likely, and trans-chelation to an endogenous protein is also a possibility for retaining radioactivity within a cell .
DNA and mRNA have been targets, but the small copy number and the challenge to design radiopharmaceuticals that demonstrate efflux of non-bound therapeutic radiopharmaceutical have prevented widespread success. Recently, Lewis et al. have targeted mRNA of bcl-2 using an antisense approach . In this case alpha and beta emitters are preferred depending on the size of the tumor. In general, the targeting of intracellular epitopes is not as developed as targeting a cell surface protein .
The emission that is most dependent on location is Auger electron emissions, which must localize in the nucleus, preferably binding to DNA, to be effective. Radioiodinated estradiol is an interesting example of transfer to the nucleus. DeSombre et al. studied two Auger electron emitters, namely, 80mBr and 125I. 2-[80mBr]Bromo-l,l-bis(4-hydroxyphenyl)phenylethylene (Br-BHPE) and the steroidal estrogen 17a-[80mBr]bromovinylestradiol both showed saturable binding to the estrogen receptor, but the decay product of 80mBr, namely, 80Br, emitted beta particles and gamma rays and therefore produced radiation dose, but lower cell killing . 80mBr is also not readily available. Next DeSombre compared [125I]iododeoxyuridine ([125I]IdU) and E-17alpha([125I]iodovinyl-11-beta-methoxyestradiol ([125I]IVME2) in vitro in MCF-7 cells expressing estrogen receptors. Both cell kill and double-strand breaks were equivalent for the two radiopharmaceuticals . The analysis of the number of receptors per cell, the achievable specific activity, and the residence time dictated that ovarian cancer metastases in the peritoneal cavity may be a more successful approach . Nevertheless, experiments in animals and theoretical calculations demonstrate the need for a large number of receptors per cell, near-theoretical specific activity of 123I, and hundreds of millicuries of 123I. Given the cost of cyclotron production of very high specific activity of 123I, human studies to affect metastases in the peritoneal cavity have not begun, but rather there is a search for a much higher-affinity constant radioligand to increase the residence time in the cell nucleus.
Antibodies and other high molecular weight proteins or particles have advantages and challenges. Many of the challenges of the size and the consequent lower permeability can be overcome by targeting proteins in the “peritoneum, pleura or intrathecal space or to radiosensitive tumors associated with the hematological system” . The challenges in designing a therapeutic radiopharmaceutical such as radiolabeled estradiol and radiolabeled antibodies are similar. The location plays a part, but the choice of radionuclide, the inertness of the bifunctional chelate, the residence time in the tumor matched with the physical half-life of the radionuclide, the scale and scope of the disease, and the absorbed dose sensitivity of the targeted tumor all come into play.
For small molecules, cell surface targets have been used most often given the access from the blood stream after IV injection. Internalization is a function of the cell as is active transport. To reach a target in the cytosol, the radiopharmaceutical must cross the cell membrane either by passive diffusion or active transport. 2-Fluoro-2-deoxyglucose enters cells via the GLUT transporter and then is metabolically trapped in the cell as the 6-phosphate. [11C]-meta-Hydroxyephedrine is taken up in myocytes by the norepinephrine receptor and metabolically trapped. To target DNA or another target in the nucleus requires transfer across another barrier. Estradiol is carried to the nucleus by several active processes.
The major questions in both cytosolic and nuclear localization are whether enough radiopharmaceutical can be delivered to the compartment and whether radiotracer not bound to the target can be removed quickly so that the detected emissions are bound radioactivity and for radionuclide therapy the decay occurs at the target.
Critical Subcellular Sites of Interaction with Ionizing Radiation (Direct Effects)
The Auger electron emitter 125I is highly cytotoxic when concentrated in the cellular nucleus as the thymidine analog 5-[125I]iodo-2′-deoxyuridine (125IUdR) . Moreover, the antitumor efficacy of Auger electron-emitting 123I or 125I conjugated to IUdR has been demonstrated in an ascites tumor model [79, 80]. Therefore, the observation that Auger electron emitters conjugated to benzylguanidine are capable of killing neuroblastoma cells raises questions about the intracellular site of concentration of MIBG and the mechanism of cell kill. Several postulates could account for these findings. MIBG may be taken up by the cell’s nucleus. It is also possible that other cellular structures are susceptible to bombardment with Auger electrons – for example, mitochondria. Alternatively, particles other than Auger electrons emitted during the decay of 123I and 125I may have sufficient range to penetrate the nucleus of a targeted cell. A fourth possibility is that MIBG labeled with Auger electron emitters mediates cell kill through apoptosis triggered by a novel mechanism.
The conventional opinion is that the critical cellular target for ionizing radiation damage is nuclear DNA. Therefore, ultrashort-range radionuclides, such as Auger electron emitters, should be toxic only if delivered to the nucleus of the target cell (Fig. 7a) . This is supported by studies in vitro – using extracellular Na125I, cytoplasmic [125I]iododihydrorhodamine, and nuclear 125IUdR, which demonstrated that significant toxicity was associated only with the 125I located in the nucleus . Similar results have also been observed in vivo by Link et al. who compared the therapeutic efficacy of 125I- and 211At-labeled methylene blue to treat mice bearing malignant melanoma . Only 211At-labeled methylene blue produced significant therapeutic effects. As the subcellular fate of this targeting agent is cytoplasmic, it was concluded that 125I-labeled methylene blue was located too far away from the genome to be cytotoxic . Although it has been shown that MIBG concentrates mainly in the cytoplasm of neuroblastoma cells [84, 85], the demonstration that significant cell kill can be achieved with 125I- or 123I-labeled MIBG could represent evidence for a nuclear localization of MIBG. The amount of drug accumulated at this site may be undetectable by conventional means but nevertheless capable of delivering a toxic dose of radiation to the cell nucleus.
Alternatively, MIBG conjugated to Auger electron emitters could be cytotoxic through damage to the mitochondria. Subcellular localization studies have identified mitochondria as sites of MIBG accumulation . Mitochondrial toxicity could result from disruption of inner mitochondrial membrane proteins or by direct effects on the mitochondrial genome. Indeed, mitochondrial DNA is particularly susceptible to radiation damage because of its limited capacity for repair .
It is also conceivable that some particles emitted during the decay of 123I and 125I may have sufficient range to reach genomic DNA of a targeted cell despite cytoplasmic or perinuclear localization. Although the entire Auger and Coster-Kronig electron spectra for these radionuclides have not been measured experimentally, calculations using theoretical transition rates and energies indicate that both isotopes emit some electrons with ranges of the order of the radius of a mammalian cell [87, 88]. Assuming a cytoplasmic location for MIBG, it is possible that such emissions would deliver significant doses of radiation to the nucleus. However, this possibility is not supported by the classical experiments of Kassis and colleagues described earlier .
A more exotic possibility is that Auger electron irradiation of cytoplasmic targets achieves cell kill by an indirect route, such as the triggering of apoptosis in susceptible cells. Some workers have reported that apoptosis-mediated cell death may result from membrane damage and the consequent generation of ceramide [89, 90] and that this process may be initiated by ionizing radiation damage to cell membranes .
Insight into the consequences of nonnuclear targeting of Auger electron s has been provided by Pouget and colleagues (see also section “Critical Subcellular Sites of Interaction with Ionizing Radiation (Indirect Effects)”). Using in vitro and in vivo models, they reported substantial antitumor efficacy of membrane-bound 125I-labeled monoclonal antibodies (125I-mAbs). Moreover, the observed cytotoxicity was greater than that associated with internalizing 125I-mAbs [92, 93].
Previous supporting experimental evidence demonstrating Auger electron-induced cell kill at a distance greater than its path length was provided by Xue et al. following treatment of human colon LS174T adenocarcinoma cells with 125IUdR .
Radiation-Induced Biological Bystander Effects (RIBBEs)
After irradiation, cells expel toxins whose consequences to neighboring, unirradiated (bystander) cells include mutation, chromosomal breakage, long-term genomic instability, and death [95, 96, 97, 98, 99, 100, 101]. Bystander effects may be induced by radiation dosage as low as 3 mGy. Such exposures are of safety concern in diagnostic imaging and intensity-modulated radiotherapy. Accordingly, it is important to understand the mechanisms underlying bystander effects and how these may be manipulated to enhance tumor cell kill.
RIBBEs are most appreciable at low radiation dose and low dose rate [102, 103, 104, 105, 106]. These are features of targeted radionuclide treatment of cancer . Therefore, bystander effects induced by radiopharmaceuticals may play a disproportionately large part in their efficacy, and understanding their nature should enable the refinement of this form of radiotherapy. The mechanisms involved are as yet undefined. However, studies using γ-ray and α-particle beams have provided some insight into possible factors. These include oxidative stress  (leading to increased radical formation [109, 110]), nitric oxide release [111, 112], cytokine release , gap junctional intracellular communication , and cellular immunity [115, 116, 117]. Agents implicated in bystander signaling have been reviewed [104, 108, 118, 119].
Although substantiation of bystander effects has been gained predominantly through studies involving external beam γ-irradiators and microbeams, some studies have been conducted of RIBBE consequent to the intracellular concentration of radiolabeled drugs. In clusters of unlabeled and [3H]thymidine-labeled cells, death of unlabeled cells was deemed to result from transmission of cytotoxins from cells which had incorporated [3H]thymidine in their DNA rather than cross fire irradiation because 3H beta-decay particles have insufficient range to impinge upon adjacent cells . Unlabeled cells were protected by dimethyl sulfoxide and lindane, implicating free radicals and gap junctional communication, respectively [121, 122]. Similarly, Xue et al. created tumors in nude mice composed of colon carcinoma cells, unlabeled or containing 125IUdR .
Again, inhibition of tumor growth was attributed to RIBBEs because the path length of 125I Auger electrons is too short for direct interaction with neighboring cells. The latter investigation demonstrated that RIBBEs are not simply artifacts of in vitro manipulations but are significant growth inhibitory factors in vivo . In an analogous study, Mamlouk et al. co-cultured [125I]IUdR-treated lymphocytes with LS174T colon adenocarcinoma cells. This induced a reduction in the rate of proliferation of the malignant cells due to the release of inhibitory factors from the irradiated lymphocytes. The absence of cross fire as a confounding element was confirmed, by the mixture of unlabeled with labeled cells, resulting in loss of bystander response . Akudugu et al. demonstrated that the capacity to succumb to 125I-induced bystander signals is not universal but is dependent on phenotype . The treatment of lymphoid cells with antibodies, labeled with the alpha and beta-particle emitter 213Bi, produced a greater than expected cytotoxicity indicative of RIBBE . This notion was supported by subsequent modeling studies of the outcome of cellular labeling with the alpha emitter 210Po .
Significantly, distinct signaling pathways have been observed to participate in the restitution of DNA in directly irradiated compared with untargeted cells. In directly irradiated cells, the kinases ataxia telangiectasia mutated (ATM), ATM and RAD3 related (ATR), and DNA-dependent protein kinase were all required for DNA restitution, whereas the activity of ATR and ATM, but not DNA-dependent protein kinase, resulted in diminished survival of bystander cells . This indicates opportunities for either sensitization or protection specifically of targeted or bystander cells or tissues through alteration of the activity of DNA damage sensors and repair enzymes.
Critical Subcellular Sites of Interaction with Ionizing Radiation (Indirect Effects)
While DNA has long been regarded as the main target of the harmful consequences of ionizing radiation, it was proposed more than 50 years ago that damage to intracellular sites other than DNA could account for the cytotoxicity of ionizing radiation . Experimental evidence indicates that the cell membrane is an important location for interaction with radiation and free radicals, resulting in the formation of cross-linkage between protein and DNA . Radiation exposure stimulates translocation from lysosomes to the cell membrane of sphingomyelinase . This catalyzes the hydrolysis of membrane-associated sphingomyelin to ceramide which, in turn, activates various kinases and transcription factors resulting in programmed cell death . The plasma membranes of cells contain combinations of glycosphingolipids and protein receptors organized in glycolipoprotein microdomains termed lipid rafts. Notably, ceramide-cholesterol complexes are responsible for the development of lipid rafts which play a significant role in cell signaling .
In contrast to the expectation that ultrashort-range, Auger electron emitters must localize in the vicinity of DNA to have cytotoxic effect, it has been demonstrated, in vitro and in vivo, that non-internalizing antibodies, directed against carcinoembryonic antigen (CEA) and labeled with 125I, have significant antitumor effectiveness. Moreover, the potency of these membrane-associated 125I-antibodies was greater than that of internalizing 125I-antibodies with specificity for human epidermal growth factor receptor (HER1). DNA damage-centered signaling pathways are effectively activated during low dose-rate Auger radioimmunotherapy . The antitumor efficacy of 125I-non-internalizing antibodies was not due to a deficiency in DNA damage recognition; nor was it induced by membrane-activated apoptosis; nor was it related to the mean absorbed dose to the nucleus. Instead, it was proposed that bystander factors were involved . These findings imply that nuclear accumulation of 125I is not necessary for its toxicity which is mediated predominantly by bystander effects independent of the subcellular site of localization (nuclear or cell membrane) of 125I.
Mitochondria make up 4–25% of cell volume, rendering them prospective targets for radiation trajectory, and it has been suggested that cells can be sensitized to radiation as a result of impaired mitochondrial function . Mitochondrial DNA is more susceptible to oxidative damage than nuclear DNA due to its proximity to the main site of generation of reactive oxygen species (ROS), the electron transport chain. Furthermore, mitochondrial DNA is not associated with protective histone proteins and has lower efficiency repair mechanisms . Reports of oncogenic products acting to maintain, in cancer cells, elevated numbers of mitochondria  which generate excessive ROS, particularly in response to hypoxia , have stimulated the development of agents that enhance mitochondrial ROS production, resulting in the apoptosis selectively of cancer cells . According to the ROS threshold hypothesis [139, 140], low levels of ROS are necessary for the maintenance of cellular metabolism, whereas excessive ROS can be cytotoxic. Therefore, the lethality of ionizing radiation and some chemotherapeutic agents is hypothesized to be due to increasing ROS concentrations to cytocidal levels preferentially in tumors which typically have elevated levels of ROS and impaired antioxidant activity as a consequence of higher rates of proliferation and metabolism . The sensitivity of cancer cells to therapeutic schemes which stimulate cell death through ROS production encourages their application in combination with the specificity afforded by targeted radiotherapy.
Significantly, mitochondria of both directly and indirectly irradiated cells are also involved in the initiation of programmed cell death  via the loss of apoptosis-inducing proteins – cytochrome C and apoptosis-inducing factor . Improved understanding of the involvement of mitochondria in the cellular response to ionizing radiation is expected to stimulate innovative targeted therapeutic strategies.
Transfectant Mosaic Spheroid Model
To adequately represent therapeutic effects upon micrometastases (a major aim of targeted radionuclides), a three-dimensional culture system is necessary. Multicellular tumor spheroids were first described by Sutherland et al. , and since then, they have been utilized for a variety of experimental applications including tumor response to various treatments [145, 146, 147, 148], basic cell biology , and investigations of microenvironmental influences on cell growth . Beyond about 300 μm diameter, spheroids from most cell lines have a well-characterized structure, comprised of a necrotic core composed of quiescent or moribund cells surrounded by a viable rim of proliferating cells. This arrangement loosely mimics the initial avascular stages of solid tumor growth in vivo, with highly proliferative activity in proximity to capillaries ranging to necrotic regions at larger distances . As a result, multicellular tumor spheroids have been utilized in a variety of experimental cancer therapy studies relevant to the treatment of micrometastases .
TMS were used in the preliminary study of a gene therapy/targeted radiotherapy strategy which conferred the potential for uptake of radiolabeled MIBG by introduction of the norepinephrine transporter gene to cells which previously did not express it . This facilitated the evaluation of the minimum percentage of transfection required to achieve cure of various sizes of metastases. A potentially beneficial substitute for 131I is the heavy halogen astatine-211 (211At) . The α-decay particles of 211At are highly cytotoxic and have a range of 55–70 μm, enhancing the prospect of tumor-selective damage. Moreover, the cytocidal efficiency of α-particles is independent of tumor hypoxia and growth rate . Boyd et al.  reported that radioactivity concentrations <20 kBq/ml of [211At]MABG induced the complete sterilization of all clonogens in 250 μm diameter mosaic spheroids composed of only 5% NAT gene transfectants. Because the path length of 211At α-particles is only 55–70 μm, cross fire irradiation from targeted to untargeted cells would be considerably less extensive than that from a β-emitter such as 131I. Therefore, this observation suggests that bystander effects, over and above cross fire irradiation, are operating in α-particle targeted therapy.
These studies indicate the potential for bystander-mediated cell kill and improved clinical efficacy of tumor targeting when only a small proportion of the tumor mass expresses the radiotherapeutic molecular target, in this case, introduced via gene modification. Moreover, RIBBE could compensate for the low levels of gene delivery currently achievable in vivo in cancer gene therapy strategies when married with targeted radionuclide therapy. Rational selection of radiohaloconjugates of MIBG will enable the enhancement of cross fire to maximize neuroblastoma cell kill.
Transfectant Mosaic Xenograft Model
To determine the toxicity of bystander signals in vivo elicited by cells which accumulated Auger electron-emitting radionuclides, Xue et al. incorporated [125I]IUdR into adenocarcinoma cells and mixed these with unlabeled cells before implantation in rodents  (see sections “Critical Subcellular Sites of Interaction with Ionizing Radiation (Direct Effects)” and “Radiation-Induced Biological Bystander Effects (RIBBEs)”). They reported inhibition of tumor growth as a consequence of factors released from the labeled cells. The transfectant mosaic xenograft (TMX) model described by Mairs et al. is an alternative means of determining the importance of bystander effects in vivo . TMX were established by subcutaneous injection into athymic mice of tumor cells comprising various proportions of norepinephrine transporter (NET) gene transfected to parental cells – a likely clinical scenario during the development of metastases of neuroendocrine tumors in vivo.
In TMX, following the administration of [131I]MIBG, retardation of tumor growth was observed in tumors comprising only 5% NET positivity. Moreover, the inhibition of TMX growth by [131I]MIBG treatment was directly related to the proportion of NET-expressing cells. TMX may be suitable for the examination of quantitative aspects of radionuclide treatment and gene therapy. For example, this model will allow, for the first time, the determination of the minimal NET expression necessary for the imaging and sterilization by radiolabeled MIBG of tumors of a range of sizes which are one to two orders of magnitude greater than TMS. The TMX model will also facilitate the determination of optimal therapeutic dosage in relation to normal tissue toxicity and the evaluation of the influence of bystander effects in targeted radiotherapy in vivo.
LET and Dose Rate
Using the media transfer methodology described in section “Experimental Models,” Boyd et al.  observed that glioma or bladder carcinoma cell lines exposed to media derived from external beam irradiated cells produced a dose-dependent reduction in survival fraction, at low dosage, followed by a plateau with respect to clonogenic cell kill at levels >2 Gy. In contrast, cells receiving media from cultures treated with [211At]MABG or [123I]MIBG exhibited dose-dependent toxicity at low dose but elimination of cytotoxicity with increasing radiation dose. Cells treated with media from [131I]MIBG demonstrated a dose-response relationship with respect to cell death and no annihilation of this effect at high radiopharmaceutical dosage. These findings suggested that bystander effect mechanisms following radiopharmaceutical administration may be LET dependent and distinct from those elicited by conventional radiotherapy.
To determine the effect of subcellular localization of radiolabeled drugs on direct and indirect cell kill, Boyd et al.  treated NET gene-transfected HCT116 cells (derived from a human colorectal carcinoma) with [131I]MIBG, [123I]MIBG, [131I]IUdR, or [123I]IUdR. [131I]MIBG and [131I]IUdR both caused a dose-related decrease in the survival of cells that had concentrated the radiopharmaceutical and in recipients of the medium from irradiated cells. The concentration of [123I]MIBG and [123I]IUdR by HCT116 cells also induced direct dose-responsive toxicity. However, bystander HCT116 cells succumbed in a dose-responsive manner to the transfer of conditioned medium from donor cultures exposed to low radioactivity concentrations of [123I]MIBG and [123I]IUdR, but the toxic effect diminished with increasing dose of radiopharmaceutical to donors. This biphasic response was similar to the previously reported effect on the survival of clonogens derived from glioma and bladder carcinoma cell lines and treated with medium from cells which had concentrated [123I]MIBG or [211At]MABG .These findings not only support previous observations of the toxicity of Auger electron emitters conjugated to compounds which do not covalently bind to DNA [76, 103, 161, 162] but also indicate that the response of bystander cells to internalized Auger electron emitters does not depend on intracellular site of concentration. Further support for this notion has been provided by Pallais et al.  who demonstrated significant inhibition of tumor growth by 125I localized to the cell membrane (see section “Critical Subcellular Sites of Interaction with Ionizing Radiation (Indirect Effects)”).
Kishikawa et al. observed, in vitro and in vivo, opposing indirect effects induced by the concentration in the DNA of adenocarcinoma cells of 123I or 125I in the form of IUdR. The bystander effect induced by 125I was growth inhibition, whereas that induced by 123I was stimulation of proliferation . These two high-LET, Auger electron emitters have similar emission spectra, but 123I (t1/2 = 13.3 h) is characterized by >100-fold greater dose rate than 125I (t1/2 = 60.5 days). Therefore, the potency of the RIBBE may be influenced not only by radiation quality but also by dose rate.
Targeting a single protein, whether on the cell membrane or in the nucleus, is necessary, but often not sufficient to produce a toxic event because of other mechanisms that are only partially understood. In targeting a specific protein, the primary factor is the location of the target, but the choice of radionuclide, the inertness of the bifunctional chelate or the stability of the covalently bound halogens, the residence time in the tumor matched with the physical half-life of the radionuclide, the scale and scope of the disease, and the absorbed dose sensitivity of the targeted tumor compared to normal tissue all come into play. But the analysis is complicated from an increasing number of observations that move toxicity from radiation at the target to the total effect. These include untargeted effects due to radiation emitted from neighboring, targeted cells as well as bystander effects produced by the cellular processing of radiation which need not impinge on DNA. Both indirect consequences of cellular radiation could make a substantial contribution to the efficacy of targeted radionuclide therapy. As more knowledge is gained of radiation-induced biological bystander effects in normal and malignant tissue, it is expected that these mechanisms will be exploited to optimize the efficacy of targeted radiotherapy and overcome the inefficiency of tumor control due to nonuniform distribution of radiation dose. The principles of molecular targeting are well established except for the paradigm shift from high-affinity, high-target density approach rather than the medium-affinity approach. Detecting lesions smaller than 2x machine resolution may favor the former approach over the latter. The design approach to take advantage of the indirect consequences of cellular radiation depends heavily on further elucidation of the indirect effect.
- 5.Adams GP. Antibody fragments produced by recombinant and proteolytic methods. In: Stigbrand T, Carlsson J, Adams GP, editors. Targeted radionuclide tumor therapy: biological aspects. Dordrecht: Springer; 2008.Google Scholar
- 24.Hillier SM, Maresca KP, Femia FJ, Marquis JC, Foss CA, Nguyen N, et al. Preclinical evaluation of novel glutamate-urea-lysine analogues that target prostate-specific membrane antigen as molecular imaging pharmaceuticals for prostate cancer. Cancer Res. 2009;69(17):6932–40.PubMedPubMedCentralCrossRefGoogle Scholar
- 25.Zechmann CM, Afshar-Oromieh A, Armor T, Stubbs JB, Mier W, Hadaschik B, et al. Radiation dosimetry and first therapy results with a 124I/131I-labeled small molecule (MIP-1095) targeting PSMA for prostate cancer therapy. Eur J Nucl Med Mol Imaging. 2014;41(7):1280–92.PubMedPubMedCentralCrossRefGoogle Scholar
- 30.Barrett HH, Alberts DS, Woolfenden JM, Liu Z, Caucci L, Hoppin JW. Quantifying and reducing uncertainties in cancer therapy. Proc SPIE Int Soc Opt Eng. 2015;21:9412.Google Scholar
- 50.Jacques Jr S, Tobes MC, Sisson JC, Baker JA, Wieland DM. Comparison of the sodium dependence of uptake of meta-iodo-benzylguanidine and norephrine into cultured bovine adrenomedullary cells. Mol Pharmacol. 1984;26:539–46.Google Scholar
- 52.Cooper MS, Ma MT, Sunassee K, Shaw KP, Williams JD, Paul RL, Donnelly PS, Blower PJ. Comparison of (64)Cu-complexing bifunctional chelators for radioimmunoconjugation: labeling efficiency, specific activity, and in vitro/in vivo stability. Bioconjug Chem. 2012;23:1029.PubMedPubMedCentralCrossRefGoogle Scholar
- 53.Szajek LP, Kao C-H K, Kiesewetter DO, Sassaman MB, Lang L, Plascjak P et al. Semi-remote production of Br-76 and preparation of high specific activity radiobrominated pharmaceuticals for PET studies. Radiochimica Acta: 2004;92, Issue 4–6, pp. 291–295. Stöcklin Memorial Issue.Google Scholar
- 69.Huclier-Markai S, Kerdjoudj R, Alliot C, Bonraisin AC, Michel N, Haddad F, et al. Optimization of reaction conditions for the radiolabeling of DOTA and DOTA-peptide with 44m/44Sc and experimental evidence of the feasibility of an in vivo PET generator. Nucl Med Biol. 2014;41:e36–43.PubMedCrossRefGoogle Scholar
- 79.Bloomer WD, Adelstein SJ. Therapeutic application of iodine-125 labeled iododeoxyuridine in an early ascites tumor model. Curr Top Radiat Res Q. 1977;12:513–25.Google Scholar
- 93.Santoro L, Boutaleb S, Garambois V, Bascoul-Mollevi C, Boudousq V. Pierre- Kotzki P-O, et al. Noninternalizing monoclonal antibodies are suitable candidates for 125I radioimmunotherapy of small-volume peritoneal carcinomatosis. J Nucl Med. 2009;50(12):2033–41.PubMedPubMedCentralCrossRefGoogle Scholar
- 146.Knuechel R, Sutherland RM. Recent developments in research with human tumour spheroids. Cancer J. 1990;3:234–43.Google Scholar
- 147.Carlsson J, Nederman T. Tumour spheroids as a model in studies of drug effects. In: Bjerkvig R, editor. Spheroid culture in cancer research. Boca Raton: CRC Press; 1992. p. 245–69.Google Scholar
- 155.Boyd M, Mairs RJ. Tumour spheroids. In: Freshney RI, editor. The culture of animal cells. 5th ed. New York: Alan R. Liss; 2006. p. 281–98.Google Scholar
- 160.Boyd M, Mairs SC, Stevenson K, Livingstone A, MCCluskey AG. Radiation quality-dependent bystander effects elicited by targeted radionuclides. J Pharm Pharmacol 2008;60:951–958.Google Scholar