1 Introduction

The history of radionuclide therapy can be traced back to the early 1900s, after the discovery of radioactivity by Henri Becquerel and Marie Curie. In 1901, Becquerel experienced some severe inflammation of his skin after placing a tube of radium in his waistcoat pocket for several hours. This discovery led to the first use of radium in therapy when Henri Alexandre Danlos and Eugene Bloch placed radium in contact with a tuberculous skin lesion. In 1903, Alexander Graham Bell suggested placing sources containing radium in or near tumours, and in 1913, Frederick Proescher published the first study on the intravenous injection of radium for therapy of various types of diseases (MacKee, 1921).

In the last twenty years, radionuclide therapy has been widely used in various clinical malignant and pain management applications (Hoefnagel, 1991; 1998; Nakabeppu and Nakajo, 1994; vande Streek et al., 1994; Forrer et al., 2006; Williams et al., 2008; Lambert et al., 2010; Nestor, 2010; Chiacchio et al., 2011; Ersahin et al., 2011; Carrasquillo et al., 2012; Ezziddin et al., 2012; Gabriel, 2012; Gulenchyn et al., 2012; Dash et al., 2013; Sainz-Esteban and Baum, 2013). Radionuclide therapy has the advantage of delivering a highly concentrated absorbed dose to the targeted tumour while sparing the surrounding normal tissues. In addition, the selective ability of radionuclide therapy is advantageous in the treatment of systemic malignancy, such as in bone metastases, where whole body irradiation using external beam radiotherapy is impossible. Since the administration of radionuclides is minimally invasive and the duration of treatment is shorter than chemotherapy, targeted radionuclide therapy has become one of the most preferred types of cancer therapy.

This paper reviews the unique characteristics and current applications of targeted radionuclide therapy, as well as future prospects including the problems and issues involved in the translation of novel targeted radionuclide therapies into clinical practice.

2 Clinical requirements and choice of radionuclide for therapy

We need to evaluate every radionuclide or radiopharmaceutical for its suitability for clinical usage. The requirements for a therapeutic radionuclide may be divided into two main categories, namely physical and biochemical characteristics (Hoefnagel, 1991). The considerations for physical characteristics include the physical half-life, type of emissions, energy of the radiation (s), daughter product (s), method of production, and radionuclide purity (Troutner, 1987; Hoefnagel, 1991; Volkert et al., 1991). The biochemical aspects include tissue targeting, retention of radioactivity in the tumour, in vivo stability, and toxicity (Zweit, 1996; Kassis and Adelstein, 2005).

The most important factor to be considered when choosing a suitable radionuclide for therapy is the effective half-life, which is the net half-life considering both physical half-life (T p) and biological half-life (T b) within the patient’s body or organs. The determination of effective half-life (T e) is explained in the medical internal radiation dosimetry (MIRD) calculation method, which is summarized as T e= T p T b/(T p+T b) (Toohey et al., 2000). The physical half-life is readily available from published radionuclide data (Weber et al., 1989), but the biological half-life requires knowledge of the radiotracer’s spatial and temporal distribution within the body. This includes radiotracer delivery, uptake, metabolism, clearance, and excretion within the patient’s body.

A suitable range of the physical half-life for therapeutic radionuclides is between 6 h and 7 d (Qaim, 2001). A very short physical half-life limits the delivery flexibility and is very impractical, while a long half-life will retain the radiation dose in the patients and expose surrounding people for a longer period. When the physical half-life is too long, patients need to be admitted and isolated for a longer period, hence increasing the treatment costs. On the other hand, the biological half-life depends on the tracer used. If the tracer is designed to be retained within the patient permanently, the physical half-life should not be too long. In addition, the tracer should have sufficient retention so that radiation can be delivered to the tumour efficiently. If the biological half-life is too short, the radionuclide will be discharged with a significantly high activity, resulting in the need for extensive radioactive waste management. Therefore, for an efficient radiation delivery, a balanced optimal biological and physical half-life should be chosen, which results in an optimal effective half-life. This consideration is best made primarily according to the type of tumour to be treated, method of administration, and uptake mechanism. If the uptake is slow, a radionuclide with a relatively long physical half-life should be used so that the radionuclide will not decay completely before it reaches the targeted tumour.

For therapeutic purposes, radiations with high-linear energy transfer (LET), such as α- and β-particles, are preferable. These types of particulate radiations allow very high ionisation per length of travel. Therefore, they are fully deposited within a small range of tissue (usually in mm). This also reduces the need for additional radiological protection as the radiation will not penetrate through the patient’s body. However, some β-emitting radionuclides also decay with γ-radiation. This associated γ-radiation could be advantageous if the energy and intensity are within the diagnostic range, as it provides the ability to visualise distribution of the radiopharmaceutical within the patient’s body using gamma scintigraphy methods. Depending on the type of tumour, the energy and intensity of the emitted radiation should be chosen so that the energy and intensity of non-penetrating radiation (i.e., α- and β-particles) are high enough compared with the penetrating radiation (i.e., γ-radiation, X-rays, or Äuger electrons), if present. Radionuclides that emit energetic α- or β-particles are preferred for the treatment of bulky tumours. However, for the treatment of small clusters of cancer cells or small tumour deposits, radionuclides that emit Äuger electrons are considered to be more beneficial because of their high-level cytotoxicity and short-range biological effectiveness (Ersahin et al., 2011).

The other factor to be considered is the daughter product of the radionuclide. An ideal radiopharmaceutical should be able to decay into a stable daughter product such as 153Sm, which fully decays into stable 153Eu. Even if the daughter product is slightly unstable, it should at least be short-lived and may eventually decay within hours into a stable product. However, if the daughter nuclide is not stable, it may contribute to the total amount of absorbed dose. Therefore, we need to consider the percentage of dose delivered from the daughter nuclide so that it will not affect the treatment plan.

For biochemical characteristics, a clinically suitable tracer should provide selective concentrations and prolonged retention in the tumour, while maintaining minimum uptake in the normal tissue (Qaim, 2001). Furthermore, depending on the mechanism of selective uptake of the tumour either by bone deposition, protein binding, or metabolic uptake, the ratio of the radiopharmaceutical’s concentration on the tumour to that of other organs or tissues should be as high as possible, so that the radiation can be delivered optimally while minimizing the radiation dose delivered to the normal tissues.

If a tracer is a proteinaceous type, it is necessary to ensure that it is stable enough during administration and is maintained in vivo long enough for sufficient radiation delivery within the patient’s body before being metabolised and eventually excreted. The other factor to be considered is the size of the tracer particles. For example, in the radioembolization of liver tumours using 90Y-microspheres, the particles should be large enough so that they will not pass through the capillary bed into other organs, especially the lungs, and at the same small enough so that they will be able to penetrate deep inside the tumour vascularity. Therefore, the typical size for 90Y radioembolization particles is between 20 and 40 µm (Houle et al., 1989; Thamboo et al., 2003; Arslan et al., 2011; Kucuk et al., 2011). Other biochemical characteristics that need to be considered include low toxicity, the specific gravity for optimal flow and distribution during administration, the appropriate pH, and the optimal clearance rate (except for permanent tracer).

Table 1 shows a summary of the physical characteristics of the commonly available therapeutic radionuclides.

Table 1 Physical characteristics of commonly available therapeutic radionuclides
Full size table

3 Selected therapeutic radionuclides and their clinical applications

Various radionuclides have been discovered or developed for therapeutic purposes since the first use of radium in the early 1900s. The most commonly used therapeutic radionuclide today is iodine-131 labelled with sodium iodide (131I-NaI) in capsule or liquid form. The treatment is widely known as radioactive iodine (RAI) therapy, which uses 131I to treat thyroid-related diseases such as Graves’ disease, solitary hyper-functioning nodule, and toxic multinodular goiter. RAI may also benefit patients with subclinical hyperthyroidism, particularly patients at risk of cardiac or systemic complications. The preferred method for treating hyperthyroidism varies in different countries. In a survey of American Thyroid Association (ATA), European Thyroid Association (ETA), and Japanese Thyroid Association (JTA) members, 69%, 22%, and 11% of respondents, respectively, named RAI as the therapy of choice for patients with Grave’s disease (Janet and Winfried, 2007). 131I has also been used for over 60 years in the treatment of patients with differentiated thyroid cancer. The efficacy of RAI therapy depends on patient preparation, tumour-specific characteristics, the disease site, and the activity of the administered radioiodine (Donald et al., 2003). The predictive value for the 10-year relapse-free survival of one negative 131I thyroid scan is about 90% after RAI therapy, whereas two consecutive negative scans have a predictive value greater than 95% (Grigsby et al., 1999).

Another major application of radionuclide therapy is in the treatment of bone metastasis. Over recent decades, several radiopharmaceutical agents have been developed and have shown evidence of safety and efficacy for pain relief in bone metastasis (Fig. 1). The radiopharmaceutical travels selectively to volumes of damaged bone, while sparing normal undamaged bone. Most of the radionuclides used are β− emitters, e.g., strontium-89 (89Sr) and samarium-153 (153Sm), which release highly energetic electrons that deposit their energy over up to several millimetres in the surrounding tissues. The treatment is effective in reducing pain from bone metastases, with pain relief reported in 60%–92% of patients in several studies (Ackery and Yardley, 1993; Lamb and Faulds, 1997; Pons et al., 1997; Giammarile et al., 2001; Anderson and Nunez, 2007). Tomblyn (2012) has reported that combining 89Sr with chemotherapy appears to improve pain response and may affect survival in patients with metastatic prostate cancer.

Fig. 1
figure 1

Serial 99m Tc methylenediphosphonate ( 99m Tc-MDP) bone scans in a patient with bone metastasis before (a) and at three (b) and six (c) months after combined treatment of 153 Sm oxabifore and denosumab (reprinted with permission from Rasulova et al. ( 2013 ))

(a) Multiple osteoblastic bone lesions were observed before the treatment. (b) The lesions showed a decreasing trend at the three-month follow-up scan. (c) Complete resolution of the osteoblastic lesions was seen at the six-month follow-up scan

Full size image

Radionuclide therapies are also used for the treatment of a variety of malignancies with 131I, 90Y, and 188Re, polycythemia with 32P, chronic joint disease with radiosynovectomy, and intravascular radiation therapy (Werner et al., 1957; Domnitz et al., 1960; Osgood, 1968; Silberstein, 1979; Hoefnagel et al., 1987; Chen et al., 1993; Deutsch et al., 1993; Parmentier and Gardet, 1994; Rhodes et al., 1996; Najean and Rain, 1997a; 1997b; Tavintharan et al., 1997; Chen et al., 2001; Buchmann et al., 2002; Clunie and Fischer, 2003; Chang et al., 2005; Schneider et al., 2005; Das, 2007; Mortazavi et al., 2007; Turkmen et al., 2007; Ugur et al., 2008; Mumtaz et al., 2009; Schultz et al., 2009; Shamim et al., 2010; de la Corte-Rodriguez et al., 2011; Kucuk et al., 2011; Lacin et al., 2011; Liepe et al., 2011; Liepe, 2012). Table 2 summarises the most commonly used radiopharmaceuticals for targeted radionuclide therapy, including their targeting mechanism and clinical indications.

Table 2 Commonly used radiopharmaceuticals for targeted radionuclide therapy (Ersahin et al., 2011)
Full size table

4 Molecular targeting radionuclide therapy (radioimmunotherapy)

Radioimmunotherapy (RIT) is a type of cancer cell targeting therapy which uses monoclonal antibodies (mAbs) labelled with a radionuclide directed against tumour-associated antigens. The ability for the antibody to specifically bind to a tumour-associated antigen increases the dose delivered to the tumour cells while decreasing the dose to normal tissues. By nature, the technique requires the tumour cells to express an antigen that is unique to the neoplasm or is inaccessible in normal cells. The appealing concept of RIT was first described by Pressman and Korngold (1953). During the subsequent decades, various RIT radiopharmaceuticals were developed with advances in genetic engineering and chelating techniques. Fig. 2 illustrates the principle of conjugation of a bifunctional chelating agent to an mAb.

Fig. 2
figure 2

Graphical illustration of the conjugation of a bifunctional chelating agent to a monoclonal antibody (adapted with permission from Milenic et al. ( 2004 ))

The chelating agent has two functionalities: one function is to bind the metallic radionuclides, and the other is to bear the reactive functional group (X) which reacts and covalently binds to N-terminal and ε-amines from lysines on the antibody

Full size image

Over the decades, an increasing number of new antibodies have been studied in clinical trials and have shown positive evidence of efficacy, particularly in non-Hodgkin’s lymphoma (NHL). The current US Food and Drug Administration (FDA)-approved radiopharmaceuticals for RIT include ibritumomabtiuxetan (Zevalin®, approved in 2002) and 131I-tositumomab regimen (Bexxar®, approved in 2003), both for the treatment of refractory NHL. Other potential applications of RIT include the treatment of breast, lung, pancreatic, stomach and ovarian carcinoma, neoplastic meningitis, leukemia, high-grade brain glioma, and metastatic colorectal cancer. Table 3 provides a summary of recently developed mAbs in advanced RIT trials and their status (Milenic et al., 2004). The clinical outcomes of these radiopharmaceuticals are discussed in the following section.

Table 3 Recently developed monoclonal antibodies (mAbs) for advanced radioimmunotherapy and the status of the clinical trials (adapted from Milenic et al. (2004))
Full size table

4.1 90Y-ibritumomab tiuxetan (Zevalin®)

90Y-ibritumomab tiuxetan (commonly known as Zevalin®) is a murine anti-CD20 antibody conjugated to linker-chelator tiuxetan. Zevalin® can be administered as outpatient therapy without significant patient restrictions. In most trials using RIT, patients with diffuse large B-cell lymphoma, follicular lymphoma and mantle cell lymphoma are recruited. In a Phase III trial by Witzig et al. (2002), RIT with Zevalin® was compared with rituximab alone in 143 patients with NHL. The overall response rates (ORRs) were 80% and 56%, and the complete response rates were 30% and 16% in the Zevalin® and rituximab groups, respectively. Alcindor and Witzig (2002) reported an ORR of 74% in patients with rituximab-refractory NHL treated with Zevalin®. Zevalin® was registered for treating relapsed or refractory, low-grade, follicular lymphoma, and transformation B-cell NHL, including refractory NHL following rituximab treatment, and approved by the FDA in 2002. In 2008, Zevalin® was approved for consolidation therapy following partial response or complete response after frontline induction chemotherapy in patients with previously untreated follicular lymphoma in the European Union.

4.2 131I-tositumomab (Bexxar®)

131I-Tositumomabor (commonly known as Bexxar®) is an mAb that is labelled with 131I. Bexxar® is used to treat certain types of NHL. Investigators have been engaged mainly in defining the role of Bexxar® RIT in treating Hodgkin’s lymphoma, diffuse large B-cell lymphoma, and multiple myeloma. Horning et al. (2005) documented that the confirmed ORR was 65% and the median progression-free survival was 24.5 months for confirmed responders after Bexxar® treatment. In a clinical trial of 60 patients with previously untreated follicular lymphoma, Bexxar® provided a significant therapeutic efficacy compared with that provided by the last qualifying chemotherapy. Based on the promising outcomes, Bexxar® was approved by the FDA for clinical practice in patients with rituximab-relapsed or refractory, low-grade follicular lymphoma in 2003 (Friedberg and Fisher, 2004). One year later, the FDA approved Bexxar® for RIT in NHL patients without previous rituximab therapy (Buchegger et al., 2006).

4.3 Epratuzumab (Lymphocide®)

Epratuzumab or Lymphocide® is a 90Y radiolabelled mAb directed against CD22 which is an antigen present on the surface of B-cells. Lymphocide® may be effective against a variety of diseases that involve abnormal B-cells, including NHL and chronic lymphocytic leukemia. It binds to cells that express the CD22 antigen on their surface and kills the cells by high-energy β− radiation from the 90Y. Lymphocide® is currently being tested in Phase III clinical trials in the treatment of lymphoma, leukaemia, and some immune diseases (e.g., lupus). The Phase I/II clinical trial results showed that Lymphocide® has a good safety profile (mostly grade 1 toxicity) and exerts anti-tumour activity in aggressive NHL at doses ≥240 mg/m2 (Leonard et al., 2004). Ongoing research includes evaluation of epratuzumab in combination with CHOP-rituximab in first line treatment of diffuse large B-cell lymphoma.

4.4 131I-Lym-1 (Oncolym®)

131I-Lym-1 or Oncolym® is an 131I-radiolabelled mAb manufactured by Peregrine Pharmaceuticals Inc., California, USA (formerly known as Techniclone). Oncolym® targets B-cells of NHL patients and is currently in Phase III clinical trials. The radiolabelled antibody will attach to a specific protein, human leukocyte antigen (HLA)-DR10 found only on the surface of B-lymphocytes, such as cancerous B-cells found in many forms of NHL. The HLA-DR10 protein is usually present on over 80% of lymphoma cells but only about 2% of normal blood cells. Therefore, Oncolym® may selectively target lymphoma cancer and spare the healthy B-cells that are required for antibody production. In Phases I and II clinical trials, DeNardo et al. (1998) found that 52% of patients that had failed conventional chemotherapy achieved partial or complete remission after receiving the treatment with Oncolym®. A possible side effect of Oncolym® is reduction of blood cell counts; however, this is usually reversible after the treatment.

4.5 131I-chTNT-1/B (Cotara®)

131I-chTNT-1/B or Cotara® is a genetically engineered chimeric human/mouse mAb that binds to the DNA-histone H1 complex, a universal non-specific intracellular antigen found in the necrotic core of solid tumours. These necrotic cells exhibit increased permeability of the cell membrane, and consequently Cotara® can successfully transfer through the cell membrane and karyolemma, target the antigen, and then deliver sufficient energy to destroy adjacent tumour cells. Recently, Cotara® was administered via convection-enhanced delivery to patients in Phase II clinical trials, the majority of whom had anaplastic astrocytoma, brain gliobastoma, hepatocellular carcinoma (HCC), or colorectal cancer treatment (Lidar et al., 2004; Patel et al., 2005; Vandergrift and Patel, 2006; Ferguson and Lesniak, 2007). In a Phase II clinical trial, 107 patients with lung cancer in eight clinical oncology centres were treated with systemic or intratumoural injections of Cotara®. The results demonstrated an ORR of 34.6%, a complete response of 3.7%, a partial response of 30.8%, no change in 55.1%, and progressive disease in 10.3% of patients (Chen et al., 2005). In 2003, Cotara® was approved for systemic or locoregional treatment of radiotherapy and chemotherapy-refractory or relapse advanced lung cancer by the China Food and Drug Administration (CFDA). A recent Phase II dose confirmation trial with Cotara® for patients with glioblastoma multiforme at the first relapse has also demonstrated a promising overall survival rate of 41 weeks (Hdeib and Sloan, 2012).

4.6 Labetuzumab (CEA-Cide)

Labetuzumab (commonly known as CEA-Cide), is a humanized mAb labelled with either 90Y or 131I that recognizes carcinoembryonic antigen (Alonso-Ruiz et al., 1998). CEA was first described in 1965 by Drs. Gold and Freedman of McGill University (Canada), as a cancer antigen associated with colorectal cancer. It has subsequently served as a serum marker and target for radiolabelled antibody imaging and other therapeutic agents because it is expressed in more than 90% of colorectal cancer cells. The Phase I/II clinical trials have shown safe and potential use of radiolabelled labetuzumab in the therapy of advanced colorectal and pancreatic cancers. In a Phase II clinical trial, 23 patients underwent Labetuzumab treatment after salvage resection of liver metastasis of colorectal cancer. The median overall survival rate was 68 months, median disease-free survival was 18 months, and five-year survival rate was 51.3% (Liersch et al., 2005). Both the median overall survival and five-year survival rates were improved compared to historical and contemporaneous controls of not receiving the labetuzumab treatment. The major adverse effect from labetuzumab was transient myelosuppression, resulting mostly in grade ≤3 neutropenia and/or thrombocytopenia.

4.7 Pemtumomab (Theragyn)

Pemtumomab (formerly known as Theragyn or HMFG1) is a mouse mAb conjugated to 90Y that binds specifically to a glycoform of the MUC1 mucin. This protein is overexpressed on the surface of epithelial tumour cells, including ovarian, gastric, breast, and lung cells. From a Phase II study in women with advanced ovarian cancer, 15 out of 21 women who were in remission at the time of treatment with pemtumomab responded well and 14 patients were still alive more than eight years after the treatment (Hird et al., 1993). Another feasible use of pemtumomab is in a gastric cancer treatment. One drawback of treatment with murine (mouse) mAbs is that it generates an immunological response in patients, known as the human anti-mouse antibody (HAMA) (Hamacher et al., 2001) response. The common responses include rashes, arthralgia, and myalgia (Maraveyas et al., 1994). It has been suggested that, in addition to any immediate effects of medication by the RIT, murine mAbs may result in a potential longer term benefit from a boost to the immune system provided by the HAMA response (Azinovic et al., 2006). If this hypothesis is correct, then the HAMA response as seen in the patients following pemtumomab treatment is encouraging.

4.8 131I-metuximab (Licartin®)

131I-metuximab or Licartin®, developed by the Fourth Military Medical University, China, is a murine mAb Hab 18 F (ab’)2 fragment against the HCC-associated antigen Hab18G/CD147, with high specificity and affinity for HCC cells. The mAb has a cytotoxic effect and potential ability to inhibit the metastasis of HCC cells by blocking fibroblast cells to produce matrix metalloproteinases. When conjugated to 131I, it can afford direct irradiation of tumour cells, and thus enhance the therapeutic efficacy. A multi-centre Phase II clinical trial by Chen et al. (2006) showed that the numbers of patients who completed two cycles of Licartin® treatment with a partial response, minor response, or stable disease were 6, 14, and 43, respectively, and the survival rate at 21 months was 44.5%. Another Chinese multi-centre clinical trial (Wu et al., 2012) showed that the survival rates at 6, 12, and 18 months were 95.6%, 80.9%, and 69.1%, respectively, in the group with combination therapy of Licartin® and transcatheter arterial chemoembolization vs. 94.3%, 72.9%, and 52.9%, respectively, in the transcatheter arterial chemoembolization (TACE) only group (Tripepi et al., 2011). In addition, the survival rate of stage II patients at one year was significantly higher than that of stage IV patients. Furthermore, the combination therapy of Licartin® and TACE could enhance the efficacy of moderate and advanced HCC patients, especially of patients with post-operation-relapse or multifocal nodular HCC (He et al., 2013). In 2006, Licartin® was approved by the CFDA for treating patients with unresectable or post-operation-relapse HCC, those who were unsuitable for TACE, and those with TACE-refractory or relapse advanced HCC.

4.9 131I-L19 (Radretumab®)

Human mAbs specific to splice isoforms of fibronectin and tenascin-C have been developed, such as L19, F8, F16, and their derivatives (Pini et al., 1998; Brack et al., 2006; Villa et al., 2008). Radretumab®, a 131I-labelled vascular tumour-targeting antibody in a small immunoprotein L19 format that binds to fibronectin, has been investigated in Phase II clinical trials (Schliemann et al., 2009). Most studies examined the role of Radretumab® in patients with non-small cell lung cancer, some solid tumours (e.g., melanoma and carcinoma of the head and neck), or haematological cancer. Results have been encouraging for the RIT of refractory Hodgkin’s lymphoma patients, and more than a hundred patients have been treated with this agent.

4.10 90Y-clivatuzumab tetraxetan (PAM4)

PAM4 is a new mAb known as clivatuzumab. It is highly reactive with pancreatic cancer and precursor lesions but is absent from normal tissues and has limited reactivity with non-pancreatic cancer. These characteristics make it a good agent for pancreatic cancer detection and therapy. In serum analysis, PAM4 has a superior sensitivity and specificity for pancreatic cancer compared to CA19-9. Recent preclinical and clinical trials showed that radiolabeled PAM4 has a great potential to be used as both an immunodiagnostic and an immunotherapeutic agent in the imaging and therapy of pancreatic adenocarcinoma (Han et al., 2014). The 90Y-labelled PAM4 is currently undergoing Phase III clinical trials.

5 Current issues and future prospects of therapeutic radionuclides

Drug delivery is an important part of targeted radionuclide therapy because merely developing an effective anticancer agent is not sufficient unless it is delivered to the site of action with an adequate dose. Conventional drug development has focused on formulations for different routes of administration, mostly oral or injectable, but this is no longer an effective strategy. Nanotechnology, advanced polymer chemistry, and biomedical engineering have contributed to the development of novel methods of drug delivery that target specific tissues without causing too much collateral damage. Innovative methods of cancer treatment, e.g., cell and gene therapies, require innovative drug delivery concepts (Panyam and Labhasetwar, 2003; Ferrari, 2005; Schmaljohann, 2006; Davis et al., 2008; Bertrand et al., 2014).

We will discuss the problems and issues that have yet to be overcome in advanced targeted radionuclide therapy and some possible improvements which can be explored for future development.

5.1 Supply issues from nuclear reactors

Various important radionuclides used for diagnostic and therapeutic purposes (e.g., 90Y, 131I, 153Sm, 166Ho, 177Lu, and 188Re) are currently produced by nuclear reactors. Although cyclotrons can produce various radionuclides and further research can result in alternative routes for several reactor-produced radionuclides, the roles of nuclear reactors have yet to be completely replaced (Qaim et al., 2011; Krijger et al., 2013). Several main reactors in the world are now aging, which poses a challenge to the constant and reliable supply of medical radionuclides globally (Perkins and Vivian, 2009). To prevent further medical radionuclide shortages, a shared global responsibility is needed by all the stakeholders to replace a small number of aging reactors with modern ones. The replacement projects should be in tandem with up-scaling projects of the most promising alternative radionuclide production routes, preferably under the auspices of an international organisation such as International Atomic Energy Agency (IAEA), to make medical radionuclides less vulnerable to shortages (Muir and Herman, 2001; Perkins et al., 2008; Pillai and Knapp, 2011).

5.2 New types of radionuclide generators

Nano-structured materials with nano-scale dimensions have given rise to a new generation of radionuclide generators. With huge surface-to-volume ratios, altered physical properties, tailored surface chemistry, favourable absorption characteristics, and enhanced surface reactivity, nano-structured materials are being explored as a new class of absorbent material in column chromatographic separation (Chakravarty and Dash, 2014). They have a potential role to play in the development of a new breed of radionuclide generators for wider applications in both diagnostic and therapeutic nuclear medicine. Nevertheless, development of such generators requires the optimal combination of a suitable nanomaterial and an appropriate parent/daughter radionuclide pair. In addition, the use of nanoparticles as effective adsorbents also requires careful synthesis, structural characterization, detailed adsorption property analysis and a reproducible scaling and manufacturing process to achieve a consistent product with the intended physicochemical characteristics (Chakravarty and Dash, 2014). This new idea is currently under development and has great potential to influence radionuclide generator technology in the future.

5.3 Contribution of molecular biology

The performance of cytotoxic drugs is defined by their selectivity of uptake and action in tumour tissue. Due to various peptide receptors being overexpressed on tumour cells, peptide receptor radionuclide therapy has been a promising therapeutic strategy that makes use of high tumour receptor expression to deliver radiation selectively to targeted tumours while sparring the normal tissues (Zaknun et al., 2013). Since radiolabelled peptides display unique characteristics, such as rapid whole-body clearance and tissue penetration, low antigenicity, and convenient production, a series of radiolabelled peptides has been designed for diagnostic and therapeutic applications in oncology (Teunissen et al., 2005; Valkema et al., 2006; Bartolomei et al., 2009; Bodei et al., 2009; Maecke and Reubi, 2011). The radiolabelled peptides have had a great impact in the management of relatively rare neuroendocrine malignancies. For example, 111In-pentetreotide (111In-DTPA-octreotide, where DTPA is the diethylene triamine pentacetate acid), which binds to the somatostatin receptors (SSTRs), has become the gold standard for the diagnosis of neuroendocrine malignancies. Radiolabelled peptide therapy is usually indicated for patients with a widespread disease that is not amenable to targeted radiation therapy or is refractory to chemotherapy (Weiner and Thakur, 2005). Phase I/II studies using various radiolabelled peptides, such as 111In-pentetreotide, 90Y-DOTA-Phe1-Tyr3-octreotide, 90Y-DOTA-lanreotide, and 177Lu-DOTA-octreotate for the treatment of neuroendocrine malignancies, are still in progress. It is known that 90Y-labelled somatostatin analogue (high-energy β− emitter) is more effective for treating larger tumours and 177Lu-labelled somatostatin analogue (low-energy β− emitter) for treating smaller tumours. de Jong et al. (2002) have suggested a combination of both radionuclides to treat patient with tumours of various sizes with non-homogeneous receptor distribution. Other peptides that are currently under investigation, some of which have shown encouraging results, include 188Re-P2045 and 90Y-αvβ3 antagonists (Weiner and Thakur, 2005).

5.4 Theranostics

Implementation of the concept of “theranostics” (also known as “theragnostics”), which refers to an integrated approach to diagnosis and therapy using suitable combinations of molecular targeting vectors and radionuclides, has started recently (Alberti, 2012). This can be achieved by using therapeutic radionuclides that also emit radiation for imaging, such as 111In, 131I, 177Lu, and 166Ho, for both therapy and diagnostic purposes. Theranostics is usually performed using molecular targeting vectors (e.g., peptides) labelled with either diagnostic or therapeutic radionuclides, targeted specifically by the vector at its molecular level (Baum and Kulkarni, 2012; Srivastava, 2012). One of the classic examples of theranostics is the use of 68Ga-labelled tracers, in which the diagnosis using 68Ga can be effectively followed by therapy using therapeutic radionuclides such as 90Y and 177Lu labelled with the same tracer for personalized radionuclide therapy (de Jong et al., 2005) (Fig. 3). The therapeutic radionuclides that are most commonly used or are expected to be widely used in the future are mainly nuclear reactor-produced radionuclides, including 47Sc, 90Y, 131I, 166Ho, 177Lu, 188Re, and 213Bi (IAEA, 2007).

Fig. 3
figure 3

Example of theranostics application

68Ga somatostatin receptor (SSTR) enables PET/CT imaging to be done for the follow-up and evaluation of molecular response to peptide receptor radionuclide therapy (PRRNT) with 177Lu-177/90Y-labelled somatostatin analogues. Molecular tumour volume (MTV) can be assessed by quantification of the SSTR density in vivo before and after the PRRNT (reprinted with permission from Baum and Kulkarni (2012)). MTI: molecular tumour index; SUV: standardized uptake value

Full size image

5.5 Contribution of nanotechnology

Radiolabelled nanoparticles that can be used as platforms for attaching different functionalities for the purposes of multi-modality molecular imaging and multi-valent targeted therapy have been proven to be promising diagnostic and therapeutic tools (El-Sayed et al., 2006; Huang et al., 2006; Gobin et al., 2007). In the last decade, nanotechnology has shown a great potential for the early detection, accurate diagnosis, and personalized treatment of various diseases, especially in cancer therapy (Sahoo et al., 2007). Their size is comparable to biological molecules such as antibodies, and about 100 to 10 000 times smaller than human cells. Nanoparticles can interact with the biomolecules both on the surface and inside the cells (Hong et al., 2009). With the capacity of large specific activity inside each particle, nanoparticles can be very useful for internal radiation therapy through passive targeting (i.e., based on the enhanced permeability and retention effect) and/or active targeting (i.e., incorporating a targeting moiety on the nanoparticle) (Mitra et al., 2006). Liposomes, spherical vesicles of lipid bilayers ranging from 100 to 800 nm diameter, are currently widely used nanoparticles for cancer therapy (Torchilin, 2005).

Other nanoparticles, which have been approved for human use or are currently in clinical trials, include iron oxide, perfluorocarbon, nanotube, quantum dot, micelle, and dendrimer (Fig. 4). The ultimate goal of nanoparticle-based radionuclide therapy is to achieve an efficient and specific in vivo delivery of therapeutic radionuclides without systemic toxicity. It will also facilitate imaging and evaluation of dose delivery and therapeutic efficacy. With the capacity to provide enormous sensitivity and flexibility, nanoparticle-based radionuclide therapy has great potential to improve cancer therapies in the near future (Hong et al., 2009).

Fig. 4
figure 4

Various types of nanoparticles that can be radiolabelled for molecular imaging and targeted radionuclide therapy (reprinted with permission from Hong et al. ( 2009 ))

Full size image

5.6 Image-based dosimetry of radionuclide therapy

Accurate dosimetric assessment during radionuclide therapy is essential to optimise treatment efficacy for the targeted sites and to minimise radiation exposure to the surrounding normal tissues (Lim, 2013). Individual dosimetry needs to be considered due to various factors contributing to the biological effectiveness in individual patients, including tissue radiosensitivity, dose rate, detailed intraorgan activity distribution, and the cycle therapy scheme (Cremonesi et al., 2006). The organ absorbed dose can be calculated based on biodistribution data derived from Monte Carlo modelling or using more accurate quantitative imaging techniques. The earlier method suggested the use of empirical fixed administered activities, or activities modified by clinical or pathological findings, for general categorized patients. This approach may lead to suboptimal under- or over-dosing as it is not individually tailored (Flux et al., 2006). Recent advances in imaging technology (e.g., quantitative single-photon emission computed tomography (SPECT) and positron emission tomography (PET) image-based dosimetry) allow precise determination of the dose absorbed to the targeted tumours and critical organs for pre-therapeutic treatment planning (individual patient-based). Image-based dosimetry prior to targeted radionuclide therapy can greatly improve the treatment efficacy and could be applied clinically in the near future.

5.7 Red marrow dosimetry

Red marrow toxicity is one of the limiting factors in many radionuclide therapies. Several studies have been carried out to develop models for red marrow dosimetry and tolerance (Stabin et al., 2002). However, red marrow has a very complex structure and the mechanisms regulating uptake activity are still unclear. The radiation burden to the red marrow varies with the radiolabelled molecules, the specific binding, and the residual activity in the blood (Cremonesi et al., 2006). The principal approaches used to evaluate red marrow dose can be categorized as blood- and imaging-based methods. The blood-based methods were specifically used in the study of radiolabelled mAbs, where there is no specific uptake in the red marrow (Sgouros, 1993). According to Sgouros et al. (1993), the activity concentration in the red marrow was linearly related to the activity concentration in the blood or in the plasma by a factor of 0.2 to 0.4, likely related to their molecular weight. The implementation of blood-based methods in peptide receptor radionuclide therapy needs an additional parameter, as peptide-bound activity in the red marrow probably distributes in a volume larger than the extracellular space. Forssell-Aronsson et al. (1995) reported a few bone marrow samples taken in patients receiving 111In-DTPA-octreotide which showed a red marrow activity concentration equal to that in plasma at the same time points. Consequently, for peptides, a factor for the concentration ratio of red marrow to blood close to one has been suggested (Cremonesi et al., 2006).

Image-based methods apply when images clearly demonstrate a specific uptake in the bone or bone marrow. Using these methods, the activity in selected areas of known red marrow volume (e.g., sacrum and lumbar vertebrae) can be quantified from the images and scaled for the whole red marrow in the patient’s body. In most studies with somatostatin analogues, no specific bone marrow localization is detected; hence, the blood activity is used as a surrogate indicator. Images of 111In- or 177Lu-labelled somatostatin analogues usually do not show any significant uptake in red marrow or in bone. Consequently, many authors prefer to extrapolate the red marrow dose from the time-activity curve in the blood, plasma, or remaining tissues (Cremonesi et al., 1999; Forster et al., 2001). The assessment of bone marrow dose still remains a challenge in most of the targeted radionuclide therapies.

5.8 RIT vs. non-radiolabelled immunotherapy

Recent advances in cancer immunotherapy, i.e., immune-checkpoint antibody blockade, adoptive T-cell therapy, anti-CD47 antibodies, and anti-GD2 antibodies, have shown a significant efficacy in the treatment of late-stage solid tumours and have become a breakthrough for cancer therapy. The anti-tumour efficacy of many therapeutic antibodies can be enhanced by their use in combination with RIT. The advantages of RIT include short treatment time and reduced side-effects. The radioactivity decays slowly, so that it is able to continue irradiating the cells over a period of time after the delivery of a single dose. Although the radiation may also kill healthy cells, the damage will be confined mostly to the adjacent targeted area. Some patients may experience nausea and vomiting, but to a lesser degree than with chemotherapy.

An earlier study performed by Witzig et al. (2002) comparing the treatment efficacy of RIT with 90Y-ibritumomab tiuxetan (Zevalin®) with a control immunotherapy, rituximab in 143 patients with relapsed or refractory low-grade, follicular, or transformed B-cell NHL, has shown that the ORR for the RIT group (n=73) was 80%, vs. 56% for the ritiuximab group (n=70) (P=0.002), whereas the complete response (CR) rates were 30% and 16% in the RIT and rituximab groups, respectively (P=0.04). The study concluded that RIT with 90Y-ibritumomab tiuxetan was well tolerated and produced statistically and clinically significantly higher ORRs and CRs compared with rituximab alone. Similar studies should be carried out in the future to explore the potential benefits of a combination therapy of radionuclide with cancer immunotherapy.

6 Conclusions

Therapeutic nuclear medicine is developing rapidly as an alternative choice of treatment in oncology. Other than its traditional uses in endocrinology and rheumatology, therapeutic radionuclides have now gained attention for tumour targeting therapy using multiple routes and mechanisms to target radionuclides in the tumour site. The non-invasiveness of the procedure and relatively low toxicity in comparison to conventional chemotherapy and external beam radiotherapy make radionuclide therapy an attractive and realistic alternative in the management of benign and malignant diseases. Although there are still many challenges, such as a low target-to-non-target rate, poorer therapeutic efficacy in solid tumours compared to haematological tumours, and a lack of radionuclide generators, we believe that, with advances in administration methods, a combination of multiple treatment modalities and enhanced therapeutic efficacy, the contribution of therapeutic radionuclides to tumour treatment will significantly increase in the future.

Compliance with ethics guidelines

Chai-Hong YEONG, Mu-hua CHENG, and Kwan-Hoong NG declare that they have no conflict of interest.

This article does not contain any studies with human or animal subjects performed by any of the authors.