Translational radionanomedicine: a clinical perspective

Introduction

Medical application has been an emerging field of nanomaterials. Nanomaterials can play a role as a multifunctional platform in medical application due to their unique characteristics as sensors and effectors. As a future medicine, nanomedicine is expected to be widely applied as in vitro nanosensor diagnostics, molecular nanoimaging or targeted nanodelivery for in vivo use and nano-tissue engineering for regenerative medicine. Among them, nanomedicine for diagnostic imaging and targeted therapy is an in vivo technology which is eventually hoped to be used in humans for various purposes. Because nanomaterials are sufficiently small to cross biological barriers including vessel walls and cell membranes, they can interact with biomaterials. Furthermore, they can be designed and modified to exhibit proper physicochemical and biological features for in vivo use. Thus, developing nanomaterials as ‘guided missiles’ which targeted specific cells and deliver drugs efficiently for diagnosis and therapy has received much attention (1). However, due to their possible toxicity and uncertainty of biological behaviors, they have limitations in human application and clinical trials are slowed down (2, 3). Even if an engineered nanomaterial has proper molecular targets and no cytotoxicity in vitro, in vivo distribution, metabolism and excretion is much more complex, which becomes a bottleneck to its entering clinical trials (4). Although more than 200 nanomedicine products have been in clinical trials or even approved (2), eventual success in clinical therapeutics is still uncertain. A feasible breakthrough for future clinical application of nanomaterials may come from radionanomedicine, an approach for clinical translation using radiolabeled nanomaterials (5–7).

The core concept of radionanomedicine lies in using ‘trace’ amounts of multifunctional radiolabeled nanoparticles for in vivo theranostic purposes. Radionanomedicine is one of the clinically feasible theragnostic approaches which combine diagnosis and therapy by a unified material to develop personalized treatment for various diseases (8). From a broad perspective, radiolabeled nanoparticles had already been used in the clinical setting as an established medical practice. For example, radiolabeled nanocolloids were used in liver scans, bone marrow scans as well as lymphoscintigraphy in nuclear medicine (9, 10). These imaging techniques rely only on the size-dependent in vivo distribution or lymphatic clearance of radionanocolloids. On the other hand, recently developed radiolabeled nanoparticles are engineered to target specific molecules of cells or to have specific capability for delivering drugs which enable both diagnosis and therapy. It is radionanomedicine in a narrow sense which includes in vivo application of nanoparticles as a multifunctional platform for monitoring in vivo behavior of nanoparticles and for looking for therapeutic effects on target tissues using therapeutic radionuclides, so-called, internal radiotherapy.

In this review, we explained how radionanomedicine would facilitate clinical application of nanoparticle-based theranostics. As a multifunctional nano-platform, recent advances of modification for radiolabeling are reviewed. In vivo behavior of nanoparticles in the light of radionanomedicine is critically reviewed. For clinical application, radiation internal dosimetry particularly for nanoparticles is discussed. As one of the clinically feasible nanomaterials, endogeneous nanoparticles are further discussed as a specific example.

Radionanomedicine facilitates clinical nanomedicine

In vivo monitoring of nanoparticles to facilitate clinical trials

Radionanomedicine can elucidate the in vivo fate of nanomaterials. Nuclear imaging of radiolabeled nanomaterials is sensitive to detect the whereabouts of administered nanomaterials in the body. The distribution study of nanomaterials can provide important information for clinical translation of nanomaterials. Firstly, in vivo monitoring elucidates whether engineered nanomaterials properly target the specific tissues such as cancer. Because nuclear imaging modalities are minimally invasive and the allowed quantification even in the deep organs in human, they can be easily translated from small animal experiments to human trials. For instance, silica nanoparticle designed for targeting αvβ3 integrin with ¹²⁴I and optical probes were experimented in small animal tumor models, which achieved tumor targeting and efficient renal clearance (19). This nanoparticles received U.S. Food and Drug Administration (FDA) approval and a clinical trial was performed, which also showed accumulation of nanoparticles in tumor lesions (20).

Secondly, positron emission tomography (PET) or single photon emission computed tomography (SPECT) provide pharmacokinetics which is a temporal pattern of the whereabouts of nanomaterials in living organisms. This pharmacokinetic information facilitates the development of clinically applicable nanomaterials. Engineering and surface modification of nanomaterials affect in vivo behavior (21, 22), thus, radionanomedicine helps to find optimal modification methods for biomedical application. The most commonly used surface modification is introduction of polyethylene glycol (PEG) to reduce opsonin protein binding in blood. This process could significantly reduce RES sequestration and increased tumor accumulation (21, 23). Size of particles as well as surface modification also affects the pharmacokinetics of nanoparticles. Small sized quantum dots are prone to be excreted by the kidneys which could result in a higher lesion-to-background accumulation ratio (24). PET and SPECT imaging could facilitate selecting clinically appropriate engineered nanoparticles which show high target lesion uptake, minimal accumulation and rapid clearance in critical normal organs.

In addition, excretion patterns and physiologic accumulation of nanomaterials explain a plausible toxicity when pharmacologic amounts of nanomaterials are used. For instance, nanomaterials quickly excreted from the body may have less systemic toxicity than those remaining in the body for a long time. A large amount of nanoparticles distributed to normal organs might cause toxicity. A nanoparticle mainly excreted by the hepatobiliary system might increase in serum concentration when hepatic functions are deteriorated (17). Therefore, in vivo kinetics of radiolabeled nanoparticles could explain possible types of toxicity by analogy with preclinical imaging.

The monitoring of in vivo kinetics of very low-dose nanoparticles could enable phase 0 clinical trials. The phase 0 trial is a platform for target modulation and proper target selection for new drug developments in human evaluating biomarkers for drug effects and pharmacokinetics (25). The phase 0 trial eventually reduces costs of the clinical trials and boosts the development. Sub-pharmacologic doses of radiolabeled drugs have been used in phase 0 trials for other drug developments which help in making an early selection of promising candidates for further clinical trials (26, 27). In particular, as biological interactions of nanoparticles are varied according to the engineering and modification of them, they are more complicated and relatively unpredictable compared with macromolecular conventional drugs. Radiolabeled nanoparticles could have an important role in phase 0 trials to facilitate rational drug selection and identify possible toxicity and therapeutic effects eventually to reduce timelines for clinical application. To accelerate phase 0 trials for radionanomedicine, pharmacokinetic modeling based on physiology of nanomaterials will be urgently needed.

Theranostic approach using radiolabeled nanoparticles

Radionanomedicine based on the target-specific delivery of radionuclides for diagnosis and therapy could be a next-generation of molecular theranostics. As radiolabeling can facilitate clinical application of nanomaterials, unique characteristics of nanomaterials may reinforce radionuclide theranostics. The principle of target-specific radionuclide theranostics is derived from nuclear medicine therapy. Classically used ¹³¹I for thyroid cancer is the best example, which enables both internal radiotherapy for thyroid cancer tissues and post-therapy scans to find residual thyroid tissues of metastasis (28). Another example is radioimmunotherapy, such as ⁹⁰Y-ibritumomab tiuxetan (Zevalin) for non-Hodgkin’s lymphoma (29). Therapeutic radionuclides can be combined with a monoclonal antibody and molecular target-specific therapy. These radiolabeled macromolecules are promising theranostic materials though some radionuclides such as ⁹⁰Y and ¹⁷⁷Lu have limitations in image-based monitoring. Another clinically available theranostic approach is radiomicrosphere therapy using ⁹⁰Y microspheres. As tumor blood supply is mostly derived from the hepatic artery, microspheres labeled with therapeutic radionuclides administered to the hepatic artery could target the tumor tissue with a high tumor-to-normal liver ratio. For a therapeutic plan as a theranostic approach, ^99mTc-labeled macroaggregated albumin is clinically used for SPECT imaging as a surrogate for the distribution of microspheres (30). Commercially available products such as resin (SIR-Spheres^®, Sirtex Medical Limited, Australia) and glass (TheraSphere^®, MDS Nordion, Ottawa, Canada) microspheres have been clinically used in hepatic metastases (31, 32) and hepatocellular carcionoma (33, 34). These clinically available radiolabeled molecules or microspheres might be extended to theranostic approaches of radiolabeled nanoparticles.

The nanoparticles could be labeled with radionuclides according to the purposes, such as for imaging, kinetics or for therapy, by only changing types of radionuclides. For instance, PEGylated liposomes with a chelator, DTPA, could be labeled with ^99mTc for SPECT imaging, ⁶⁸Ga for PET imaging and ¹⁷⁷Lu for therapeutic purposes (35). Therapeutic radiolabeled nanoparticles could be not only intravenously injected but also directly delivered to target lesions. The image-guided theranostic approach using radiolabeled nanoparticles achieved by intratumoral administration resulted in high radiation dose delivery to the tumor tissues (36). Furthermore, simultaneous imaging and delivery could be an example of theranostic approach of radiolabeled nanoparticles as nanoparticles loaded drugs rather than therapeutic radionuclides. Nanocarriers for simultaneous drug delivery and imaging such as biocompatible mesoporous silica labeled with ⁶⁴Cu could be applied to a image-guided drug delivery system (37). Recently developed nanocarriers could selectively release drugs or be activated in accordance with external environments for effective treatment. Such system using a gold nanorod as a multifunctional platform could be labeled with PET agents to realize a combination of chemotherapy, photothermal therapy and PET imaging (38).

Recent evidence of tumor heterogeneity implied that we might need multiple molecular targets and ligands for successful cancer therapy (39, 40). Nanomaterials can provide multifunctional platforms for radiolabeling which enables simultaneous imaging and therapy against the multiplexed targets for proper personalized combined internal radiotherapy. Using the recent one-step surface modification for nanomaterials to have multiple functions, nanomaterials can simultaneously possess multiple targeting ligands and multiple radioisotopes labeled for diagnostic and therapeutic purposes (7, 41). The multifunctionality and flexibility of radiolabeled nanomaterials will promote the radionuclide theranostics.

Approaches for clinical radionanomedicine

Modification strategy for clinical application

Most nanoparticles in biomedical research have been delivered to the specific targets of cancer and inflammation. The accumulation of nanoparticles was dependent on two different mechanisms, ‘passive targeting’ and ‘active targeting’. Passive targeting is a result of enhanced permeability and retention (EPR). Leakage in inflammation or cancer tissues facilitates the accumulation of nano-sized materials in these lesions (56). Though nanoparticles accumulate effectively in most cancer tissues according to this EPR effect, passive targeting is limited to assure the delivery of therapeutic nanoparticles because tumor is heterogeneous and contains hypoperfused regions where nanoparticles can hardly reach and be accumulated (57). Theranostic nanoparticles have been engineered to have multiplexed targeting ligands on their surfaces to enable active targeting. Most commonly used bioactive molecules are monoclonal antibodies (58). However, antibodies are limited because the relatively large size of antibodies shall increase the diameters of nanoparticles and also the immunogenicity of these nanoparticles (34). Antigen binding proteins, part of monoclonal antibody, and single chain fragments are preferred for surface modification for active targeting due to the higher targeting efficacy with lower immunogenicity (59). And recently, affibodies, aptamers, engineered peptides and small proteins are favored for active targeting. Small molecules are also the favored candidates as the ligands for active targeting because of their ease of conjugation and low immunogenicity (60). Despite the active process of targeting, the real specificity of the uptake of nanoparticles mainly depends on the bioavailability in addition to the ligand-target specificity. As it is difficult for most nanoparticles to escape from the opsonization and the phagocytosis, which is discussed in the later part of this review, nanoparticles should be modified to increase bioavailability for successful active targeting (61).

To overcome the complexity of surface modification and radiolabeling of nanoparticles, we need to develop simple and appropriate procedures. Nanoparticles can be designed for active targeting, radiolabeled for imaging and modified for increased bioavailability. Because of diverse types of modification for nanoparticles, in vivo behavior of them should be tested and favorable candidates would be selected for further clinical translation. As an example of nanoparticles for tumor targeting, nanoparticles should be highly accumulated in the tumor and rapidly excreted from the other organs to reduce the possible toxicity. Nanocarrier systems should effectively deliver and release drugs to the target area (37). Because in vivo kinetics and targeting properties according to such process can be visualized by radiolabeling, various types of engineering such as different active targeting ligands (e.g. RGD peptides) should be tested. Recently, a quick and simple modification method to achieve multifunctionality was developed (41). ‘Jeong’s method’ can make the most metallic or metal-oxide nanoparticles with hydrophobic surface to hydrophilic and give the multifunctionality in one-step (Figure 1). The beauty of Jeong’s method is the simplicity of the procedure, i.e. mixing of the functionalized micelle and hydrophobic nanoparticles followed by gel filtration after vortexing. The total modification time would be <2 h including size exclusion chromatography. In addition, the functionalized nanoparticles modified by this approach can have a monodispersed size-distribution even after the whole modification and radiolabeling step (62). Such a process helps in vivo tests of nanoparticles to select candidates for further clinical translation which show a high lesion-to-normal organ concentration ratio and appropriate targets. We expect the convenience and robustness of this approach will contribute and boost the translational research of nanomedicine.

Figure 1:

The one-step method for multifunctional nanoparticle.

Clinical translation of radionanomedicine needs quick and simple methods for modification of nanoparticles. Radiolabeled multifunctional nanoparticles can be prepared by simply mixing functionalized micelles and hydrophobic nanoparticles. The modification is followed by radiolabeling. This approach facilitates clinical application as various modified nanoparticles can be tested in vivo and proper candidates for further clinical trial can be selected.

Issues of in vivo behavior investigated by radionanomedicine

A key advantage of radionanomedicine lies in its capability to quantitatively examine in vivo distribution of nanoparticles. In vivo distribution assessed by PET or SPECT imaging provides crucial information of the where-abouts of nanoparticles, prediction of toxicity, dosimetry for specific tissues and excretion patterns (63). Fluorescence imaging of nanoparticles themselves (e.g. quantum dots) or labeled dyes onto the nanoparticles is an easy and simple way to detect in vivo distribution of these nanoparticles. However, this optical imaging is not suitable for human application due to the limited penetration depth as well as the inherent toxicity of the nanoparticles. Serial imaging is possible with radiolabeled nanoparticles even for the organs deep in humans, which allows further quantitative analysis. In spite of the strength of radiolabeled nanoparticles for in vivo tracking, there are several potential pitfalls for clinical translation of radiolabeled nanoparticles.

In vivo distribution and targeting properties are affected by physicochemical factors of radiolabeled nanoparticles. The factors controlling kinetics of nanoparticles were reviewed comprehensively by Moghimi et al. (64). Size and shape of nanoparticles are major factors which influence the clearance, phagocytosis and duration of circulation (64, 65). Particle geometry and surface charge affect the cellular internalization mediated by endocytosis. For instance, non-spherically shaped and positively charged particles let the nanoparticles to be internalized easily (66, 67). Several biological factors as well as these physicochemical properties also affect in vivo distribution. The influence of biological factors on the distribution pattern is so complex that animal studies hardly warrant immediate clinical translation. Several biochemical reactions take place following in vivo administration of nanoparticles, which differ between animals and humans. This is also the case with radiolabeled nanoparticles. Once nanoparticles are administered in body fluids, they will encounter various molecules in our body, which alter the surface of nanoparticles. These biologically-generated collections of proteins on their surfaces are called protein corona. Their composition and patterns change according to the physical characteristics of nanoparticles as well as the nature of human body fluids (68–70). Protein corona are known to significantly affect in vivo distribution and targeted delivery (71, 72). Moreover, the composition of protein corona changes according to the disease status and even shows person-to-person difference (73, 74), which makes inference more difficult about clinical translatability of in vivo application of nanoparticles based on small animal experiments.

Another difficulty in clinical translation of in vivo behavior revealed by radionanomedicine is the fact that radiolabeled nanoparticles are only a small fraction of nanoparticles. Physical properties of nanoparticles rely on preparation steps, which largely affect in vivo behavior of nanoparticles. Even more, size and electrical properties of nanoparticles can be different when the samples are prepared with the same procedures (75). In addition, distribution of properties of nanoparticles means that samples of nanoparticles are inhomogeneous to a certain extent. Radiolabeling procedures could affect the variability of properties of nanoparticles because simple procedures can make aggregated nanoparticles (76). Furthermore, radiolabeled nanoparticles could be slightly different from unlabeled nanoparticles because radiolabelling procedures could alter the nanoparticles to some extent. Nanoparticles with chelators could affect surface areas and shapes of nanoparticles which alter the in vivo behavior compared with unlabeled nanoparticles (77). Characterization procedures of nanoparticles such as electron microscopy, dynamic light scattering and zeta-potential measurements should be critically performed for radiolabeled nanoparticle compared with unlabeled one for clinical translation.

Degradation and metabolism of radiolabeled nanoparticles in vivo should be considered for clinical translation. The serum stability of radiolabeled nanoparticles is confirmed in vitro before in vivo experiments in most cases, however, their clearance in vivo further depends on their interaction with the in vivo biological milieu (59, 78). Clearance from the body of the nanomaterials mostly involves metabolism in the liver followed by intestinal excretion or independent renal excretion. Systemically administered nanoparticles are regarded as foreign materials by the bodily defense system and coated by opsonin proteins. This process is followed by phagocytosis particularly in the mononuclear phagocytic system such as Kupffer cells in liver and macrophages in the spleen and lungs (79). To reduce the opsonization, hydrophilic coating with polyethylene glycol (PEG) has been tried as a popular surface modification. However, opsonization takes also place with these hydrophilized PEGylated nanoparticles. Though the inorganic nanoparticles coated with organic compounds are generally stable in vitro, they could easily be degraded in vivo (80). This degradation of radiolabeled nanoparticles makes in vivo distribution of nanoparticles more unpredictable. If the degradation of radiolabeled nanoparticles occurs in vivo before they reach the target, nuclear imaging might yield incorrect information regarding the in vivo distribution of nanoparticles (80). To clarify these issues, distribution and excretion of radioisotopes themselves or the radioisotope-chelator complex should be studied simultaneously with radiolabeled nanoparticles in preclinical models (5, 81). Figure 2 summarizes the pitfalls of interpretation of the distribution of systemically administered radiolabeled nanoparticles using radioactivity detection.

Figure 2:

Pitfalls in tracking radiolabeled nanoparticles.

Systemically administered nanoparticles encounter various biological molecules and several reactions take place. Protein corona is generated around the nanoparticles and opsonized. After this immediate reaction, injected nanoparticles could induce phagocytosis particularly in mononuclear phagocyte systems. Radiolabeled nanoparticles could be degraded and nuclear imaging might delineate only the radioactivity and yield inaccurately the distribution of nanoparticles.

Excretion and degradation of systemically administered nanoparticles are also complex. The excretion and degradation of nanoparticles are generally mediated by opsonization and phagocytosis. After these clearance processes, nanoparticles and their metabolites are excreted via hepatobiliary or urinary tract (82). The excretion of multiple metabolites of engineered nanoparticles should take place via both routes, however, nanoparticle excretion is varied between phagocytosis, hepatobiliary and/or renal excretion according the characteristics of each radiolabeled nanoparticle. In an example of a radiolabeled upconversion nanoparticle (UCNP), radiolabeled UCNPs gathered in hepatic vascular sinusoids, were taken up by hepatocytes and finally excreted via bile canaliculi in their integral form rather than been taken up in Kupffer cells (83). This study of the excretion pathway using radioactivity and upconversion-luminescence imaging using ⁶⁴Cu-labeled UCNP confirmed the excretion in the integral form. This was different from excretion of another radiolabeled nanoparticle after the capsular detachment from the nanoparticles (83). As a different excretion pathway, direct excretion from intestines mediated by goblet cells was also reported, not mediated by hepatobiliary excretion (84). Due to this complexity of in vivo distribution and excretion of nanoparticles which might even vary between individuals according to the types of nanoparticles, an individualized distribution study is required to consider the future clinical application.

To sum up, in vivo distribution is affected by several individual factors as well as physical characteristics of nanoparticles. In vitro stability cannot be translated directly to in vivo situations because of various biochemical reactions occur immediately after administration. In addition, excretion and metabolism of radiolabeled nanoparticles make biodistribution study harder to interpret. The complexity and uncertainty of distribution, metabolism and excretion of nanoparticles suggest that all nanoparticles need individualized human biodistribution study based on radionanomedicine.

Dosimetry of radionanomedicine

Radiation dosimetry is crucial for the clinical use of radiolabeled nanoparticles for therapy as well as diagnosis. Dosimetry data are required for radiation safety as well as measurement of effective radiation doses for internal radiotherapy. Radiation doses absorbed in specific tissues are determined by kinetics of radiopharmaceuticals, physical characteristics of radionuclides such as half-life, emission types and retention time (85). To calculate radiation doses to the specific target organs, total cumulated radioactivity calculated by integration of time radioactivity curve is obtained first. Absorbed radiation doses to the target organs consist of radiation from the radioactivity of target organs themselves and radiation from the neighboring organs. This calculation process is standardized in the open-source softwares such as OLINDA/EXM (86). Pharmacokinetics of radiolabeled nanoparticles and anatomical distribution are key points to calculate dosimetry.

Although radiation dosimetry of radiolabeled nanoparticles is the essential for clinical application, few preclinical studies have been reported. Preclinical dosimetry study using therapeutic radiolabeled nanoparticles can provide the information for basic treatment planning with regard to the administration dose for successful clinical trials. Delivery of the highest absorbed dose to the target tissues satisfying safety dose which limits to the critical organs including bone marrow, liver and kidneys are the key issues in designing therapeutic application of radiolabeled nanoparticles. They depend upon the types of radionuclides given and their pharmacokinetics varying with modified surface characteristics. In preclinical studies, dosimetry using radiolabeled liposomes was performed to determine which types of liposomes were the most effective for delivering optimal dose and least radiation to bone marrow (87–89). Furthermore, radiation dosimetry of nanoparticles for diagnostic imaging should be performed to predict safety in humans. In a recent report of first-in-human study using ¹²⁴I-labeled silica nanoparticles (20), radiation doses were reported as the fundamental safety data for further clinical trial. Because radioiodine-labeled nanoparticles increase concerns of free iodine-related thyroid damage, this study focused on radiation doses to the thyroid as a critical organ.

Radiation dosimetry has been routinely performed in developments of new radiopharmaceutics. It should be also performed before clinical application of radiolabeled nanoparticles. Even though radiation dosimetry of nanoparticles is almost similar to that of conventional radiopharmaceutics, there are some points that need to be carefully considered. Firstly, nanoparticles for internal radiotherapy can be labeled with multiple radionuclides, which can make it harder to estimate radiation dosimetry. Recent reports suggest that the combination of different types of beta-emitters could be effective for internal radiotherapy according to the tumor size and uptake of nanoparticles (90, 91). Various therapeutic radionuclides with different beta-ray energy such as ⁹⁰Y and ¹⁷⁷Lu can be combined for a personalized optimal ratio according to the tumor properties (90). Therefore, radiation dosimetry of multiple radionuclides might be considered the future for radiolabeled nanoparticles particularly for internal radiation therapy to tumors. In addition, because of multifunctionality, radiolabeled nanoparticles can deliver chemical drugs as well as therapeutic radionuclides. In vivo kinetics obtained by radiolabeled nanoparticles can provide radiation dosimetry as well as estimated concentration of chemical drugs in the target tissue. For example, ¹⁸⁶re-labeled liposomal doxorubicin could be applied to simultaneous local radiation and chemotherapy for tumor tissues (92). Dosimetry calculated by accumulation of the radiolabeled liposomes helps determine both radiation and chemotherapy doses for the treatment.

In the future, every clinical trial using radiolabeled nanoparticles should be accompanied by appropriate dosimetry studies. Considering that the internal dosimetry is done by using the kinetic data of radiotracer administered in vivo, the preclinical kinetic modeling and the precise quantification of radiolabeled nanoparticles must precede the ensuing dose-optimization studies.

Radionanomedicine of endogenous nanoparticles

Inorganic exogenous nanoparticles have been studied for biomedical application with the due concerns about in vivo safety. As we do not use a pharmacologic amount but a trace amount of radiolabeled nanoparticles, we are not worried about the possible toxicity of these nanoparticles. Nevertheless, we still need to care about substantial biocompatibility of these nanoparticles for facilitating clinical nanomedicine. In this aspect, endogenous nanoparticles recently received great attention. Most living things ranging from unicellular organisms to complex multicellular organisms exploit extracellular vesicles for intercellular communication (93). We can consider these extracellular vesicles as endogeneous nanomaterials, and in contrast, liposomes can be considered as exogenous organic nanomaterials which are extracellular vesicle-mimetic. Because they can be engineered for diagnostic and therapeutic purposes as exogenous nanomaterials, we expect that we could avoid the concerns about toxicity and possibly the immune responses and thus facilitated clinical application is expected.

Extracellular vesicles are a broad spectrum of vesicles excreted by various types of cells. They can be found in most types of body fluids including serum, urine and saliva. Recently, as biological nanomaterials, they have been used as emerging therapeutics for oncology (94, 95) and regenerative medicine (96). Extracellular vesicles basically consist of a lipid bilayer and contain the proteins, the cytosolic components and the nucleic acids which carry biological information for intercellular communication (97). Thus, they can be used as nanocarriers which transport genetic materials and proteins as well as drugs (94, 98). Simple loading ability of both lipophilic drugs (doxolubicin or curcumin) and hydrophilic molecules (therapeutic nucleic acids) highlights the potential attributes of exosome (extracellular vesicles) as delivery vehicle. Another example of therapeutic use of extracelluar vesicles for cancer therapy was exosome-based cancer immunotherapy. Given that exosomes deliver tumor-antigens to elicit immune responses, exosome-derived anti-tumor immune response was proposed as cancer therapeutics. Exosome-based cancer immunotherapy derived from dendritic cells was already performed as a phase I clinical trial (99–101).

Even though biomedical application of extracellular vesicles is promising, in vivo behavior and distribution of administered extracellular vesicles are still unknown. In spite of advantages in biocompatibility of extracellular vesicles for in vivo administration, in vivo kinetics and behavior have not been fully understood. Thus, radiolabeled extracellular vesicles for quantification of systemic distribution could facilitate clinical translation of them because they have less concern about safety than other nanoparticles (102, 103). Recently, radiolabeled extracellular vesicles disclosed biodistribution of systemically administered extracellular vesicles, either natural exosomes or exosome-mimetic nanovesicles. Using reporter systems expressed in exosomes to have membrane-bound streptavidin, ¹²⁵I was successfully labeled to exosomes by conjugation of norbiotinamide and biodistribution could be elucidated (104). ¹¹¹In-oxine (105) or ^99mTc-HMPAO (76) was used for radiolabeling which used the principle of entrapment of radiotracers converted in the cytosol from hydrophobic to hydrophilic compounds. These radiolabeling with entrapment could be a biocompatible option as the entrapment procedures could be performed in physiologic condition only mixing extracellular vesicles with the radiotracers. SPECT images were successfully acquired in animals using ^99mTc-HMPAO-labeled extracellular vesicles (76). So far, most studies of exosomes presumed that they have target-specific homing effect (106, 107). Several studies which used exosomes for therapeutics focused on therapeutic effects on target tissues without in vivo distribution study. The radiolabeled exosomes can elucidate whether they target the specific tissues as well as disclose the distribution, i.e. physiologic behaviors of exosomes in the body.

For clinical application of extracellular vesicles of radionanomedicine, a lot of further studies are warranted. Firstly, low yield of exosomes produced from cultured cells limits the evaluation of therapeutic usability of extracellular vesicles as well as imaging studies with radiolabeling. Vesicles secreted from cells are isolated and purified by various methods including ultracentrifugation, filtration and immunoaffinity isolation from culture media or biological fluids (108). However, large-scale production for in vivo application is difficult. A recently developed method which collects exosome-mimetic nanovesicles from whole cells can be an alternative (109). Radionanomedicine with high detection sensitivity can contribute to overcome this problem of low yield of radiolabeled exosomes. Secondly, just as exogeneous nanoparticles are metabolized and degraded in the body, extracellular vesicles are also degraded rapidly (104, 110). Thus, whereabouts of radionuclides might not reflect the accurate distribution of the vesicles themselves. Simple measurement by thin layer chromatography capable of distinguishing the radiolabeled extracellular vesicles from the degraded extracellular vesicles would be helpful to better understand in vivo stability and distribution of the radiolabeled extracellular vesicles. Furthermore, as extracellular vesicles are derived from a variety of cells or found in any body fluids, their composition also varies a lot according to the condition of the cells (111). Recent evidence on in vivo behavior of extracellular vesicles clearly demonstrated that their in vivo distribution varies depending on the cell sources or the administration routes (112). For examples, bone marrow dendritic cell-derived exosomes were inclined to splenic tissue, but muscle cell-derived exosomes or melanoma-derived exosomes accumulated more in the liver and the lungs, respectively. Their differential intrinsic homing ability will help guide the design of in vivo therapeutic extracellular vesicles as carriers to deliver the drug to the tissue of interest. Interestingly, the same group documented there were no differences of cross-species distribution of extracellular vesicles between human and mouse-originated extracellular vesicles in a mouse xenograft model despite different cell origin-dependent tissue tropism (112). This complexity and diversity of extracellular vesicles suggest that the imaging of radiolabeled extracellular vesicles should be routinely performed if they are further considered for their usefulness for clinical theranostics.

Conclusion: future perspectives of radionanomedicine

Targeted delivery of nanoparticles for theranostic purpose is promising for future clinical application due to their multifunctionality and flexibility. However, the clinical translation is limited due to the concerns of toxicity and uncertainty of in vivo physiology. Radionanomedicine provides in vivo behavior of nanoparticles using traceable amounts of nanomaterials which would avoid any predicted pharmacologic effects. In principle, this principle of radiotracer technology and noninvasive nuclear imaging could facilitate clinical trials. Furthermore, using therapeutic radioisotopes, internal radiotherapy based on radiation dosimetry can be achieved with image-based monitoring.

For clinical translation of nanoparticles initiated from radionanomedicine, several issues related with radiation and interpretation of results should be carefully considered. In particular, in vivo behavior of radiolabeled nanoparticles is too complex to be simply interpreted by tracer amounts of them, because it is affected by physicochemical and biological characteristics of nanoparticles. Variability of nanoparticles and alteration in a biological milieu that might be associated with individual-to-individual difference make it harder to interpret biological behavior of nanoparticles. Therefore, development of a procedure of simple and appropriate radiolabeling and surface modification is urgently warranted because demands are great for the characterization of in vivo behaviors of various types of nanoparticles. The fact that in vivo distribution and excretion of nanoparticles are varied between individuals suggests that clinically applicable nanoparticles should be routinely examined for their biodistribution using radionanomedicine discipline. In addition, for the clinical application of radiolabeled nanoparticles, internal radiation dosimetry is required as an essential safety assurance using quantitation of expected dose for internal radiotherapy. As the dosimetry is also based on the kinetics of radiolabeled nanoparticles, kinetic modeling and distribution revealing the contribution of physicochemical and biological aspects of nanoparticles should be comprehensively studied. The periodic table of nanoparticle physiology according to the types of nanoparticles in various biological milieu will eventually be constructed using radiolabeled nanoparticles (5).