Bifunctional chelators for radiorhenium: past, present and future outlook

Abstract

Targeted radionuclide therapy (TRNT) is an ever-expanding field of nuclear medicine that provides a personalised approach to cancer treatment while limiting toxicity to normal tissues. It involves the radiolabelling of a biological targeting vector with an appropriate therapeutic radionuclide, often facilitated by the use of a bifunctional chelator (BFC) to stably link the two entities. The radioisotopes of rhenium, ¹⁸⁶Re (t_1/2 = 90 h, 1.07 MeV β⁻, 137 keV γ (9%)) and ¹⁸⁸Re (t_1/2 = 16.9 h, 2.12 MeV β⁻, 155 keV γ (15%)), are particularly attractive for radiotherapy because of their convenient and high-abundance β⁻-particle emissions as well as their imageable γ-emissions and chemical similarity to technetium. As a transition metal element with multiple oxidation states and coordination numbers accessible for complexation, there is great opportunity available when it comes to developing novel BFCs for rhenium. The purpose of this review is to provide a recap on some of the past successes and failings, as well as show some more current efforts in the design of BFCs for ^186/188Re. Future use of these radionuclides for radiotherapy depends on their cost-effective availability and this will also be discussed. Finally, bioconjugation strategies for radiolabelling biomolecules with ^186/188Re will be touched upon.

This review highlights some of the past and current bifunctional chelators developed for radiorhenium (¹⁸⁶Re and ¹⁸⁸Re), as well as providing an outlook on what we may expect in the field of rhenium radiopharmaceuticals in the future.

Introduction

As the second leading cause of death globally, cancer continues to be a major public health issue and is currently responsible for 1 out of every 6 deaths recorded worldwide.¹ Of particular concern is the fact that approximately 70% of these deaths occur in low- to middle-income countries, placing great stress on an already weak economic and health care infrastructure.^1,2 Currently, the main forms of non-surgical treatments across all cancer types include chemotherapy or external beam radiotherapy (EBR), both of which have proven to be only partially successful in prolonging the lives of cancer patients.³ The main drawback for these treatments is their poor selectivity for cancer cells, with many rapidly-dividing, nontumour cells (such as blood cells and those in the gut and hair follicles) being damaged or destroyed over the course of the treatment.⁴ In fact, the clinical trial success rate of new chemotherapeutic agents in oncology is only around 3%,⁵ partially due to the fact that the drugs undergoing the clinical trial investigations often end up targeting the wrong pathway, leading to off-target toxicity.^6,7 The reason for this failure in cancer drug efficacy is not very well understood, but is the driving force behind the search for more targeted therapies which are designed to target solely cancerous cells while sparing healthy tissue.

In recent years, targeted radionuclide therapy (TRNT) has emerged as a promising and rapidly proliferating approach to treat many types of malignancies.^3,8,9 In a similar fashion to the “Magic Bullet” concept of targeted therapy described by German physicist, Paul Ehrlich, TRNT is able to locate a specific tumour-associated target and deliver radioactive atoms to destroy it.^10,11 While standard chemotherapies commonly act indiscriminately on both normal and cancerous rapidly dividing cells, TRNT brings the radionuclide directly to the tumour via a specific, cell-targeting vector. Once the radiopharmaceutical is retained at the site of action, the radiation that it produces can then kill the vicinal cancer cells.^12,13 Although some radionuclides are non-metallic and can be incorporated into the targeting molecule, most therapeutic radionuclides are metal-based and therefore require the use of a bifunctional chelator (BFC). A BFC is a molecule which has both a metal-binding moiety (i.e. the chelator) to strongly coordinate a radioactive metal ion and a chemically reactive functional group suitable for the covalent attachment of a targeting vector (Fig. 1).^14–16 Depending on the oxidation state of the radiometal, BFCs are designed bearing specific donor atoms and frameworks to ensure the formation of a radiometal chelate that has both high thermodynamic stability and kinetic inertness.^16,17 The radiometals used in TRNT are known as radionuclides, the choice of which is primarily dependent on the eventual application (diagnostic imaging or therapeutic) of the radiopharmaceutical. Deeply-penetrating, γ- and β⁺-emitting radionuclides are mostly used in diagnostic imaging while the choice of radionuclide for therapeutic applications is highly dependent on the type, size and location of the tumour.³ Therapeutic radionuclides mostly emit either α-particles, β⁻-particles or Auger electrons, some of which have much shorter path lengths than those used in imaging.¹⁸ In addition to their good physical and chemical characteristics for radiotherapy, β-radiating radionuclides are more cheaply produced and widely available than other types of particle emitters, making β-therapy the most feasible and frequently used emission type for TRNT.^3,7,19 In the context of anti-cancer therapy, there are currently thirteen β-particle-emitting radionuclides used in both clinical practice and in pre-clinical studies.³ Of these, the ¹⁸⁶Re and ¹⁸⁸Re isotopes have emerged as two of the more promising radioisotopes for use in TRNT. Both ¹⁸⁶Re and ¹⁸⁸Re possess favourable physical characteristics and nuclear properties which offer a myriad of advantages for potential therapeutic applications.^20–22 These will be discussed later on in this review.

Fig. 1. The bifunctional chelator approach for TRNT.

Over the past decade, many interesting and comprehensive reviews on the different types of BFCs for radiorhenium in the context of TRNT have been written.^{13,18,22–25} In fact, in 2017 an entire special issue by the International Journal of Nuclear Medicine and Research was dedicated to the use of ¹⁸⁸Re in radionuclide therapy.²⁶ The purpose of this review is to provide a brief summary of some of the most notable chelators, as well as give an up-to-date account of the BFCs developed for ¹⁸⁶Re and ¹⁸⁸Re over the past few years. It will furthermore examine the successes and failings of radiorhenium within TRNT as a whole. Considering its history until now, we also aim at providing a future outlook for the application of radiorhenium in TRNT in the advancing modern age of nuclear medicine.

Why radiorhenium?

Named after the Rhine river in Europe, rhenium is a fascinating metal with a rich chemistry. Although complexes of rhenium have long been overlooked for their use in anticancer therapy, recent literature shows they offer great potential as potent anticancer agents in the effort to overcome the clinical limitations observed with platinum-based drugs.^9,27–31 Rhenium merits particular attention among all other transition metals because of its wide range of oxidation states ranging from −1 to +7 and coordination numbers up to 9.^16,32,33 This allows it to form structurally diverse complexes with a variety of ligands and BFCs that would not be possible with other transition metals. Comparatively, platinum, the metal used in the well-known anticancer drug, cisplatin, only has two predominant oxidation states (+2 and +4).³³ As a more versatile metal with a different range of mechanisms of action, rhenium thus proves to have a number of advantages over platinum in the design of novel chemotherapeutics.³³

Rhenium has two medically relevant isotopes (¹⁸⁶Re and ¹⁸⁸Re), the nuclear properties of which are summarised in Table 1.^{20,23,33–35} They are β⁻-particle emitting isotopes which each have different half-lives and particle energies, making them suitable for different tissue ranges in the body.^34–36 The shorter tissue penetration range of ¹⁸⁶Re means it is suitable for the targeting of small tumours while ¹⁸⁸Re is appropriate for much larger, heterogeneous masses for which deep tissue penetration is required.^32,34,35 Additionally, both rhenium isotopes also co-emit γ-radiation, which allows for in vivo imaging of ^186/188Re-labelled biomolecules as well as for dosimetry calculations.^13,34

Nuclear properties of the rhenium radionuclides, ¹⁸⁶Re and ¹⁸⁸Re.

Radionuclide	t 1/2 (h)	β−Emax (MeV)	γ Emax (keV) [%]	Tissue penetration range (mm)
186Re	90	1.07	137.2 [9.47]	4.5
188Re	17	2.12	155 [15.6]	11

The choice of rhenium radionuclide often comes down to factors such as the half-life (t_1/2) and the ease of radioisotope production. ¹⁸⁶Re has a half-life of roughly 3.7 days, thus providing sufficient time for the synthesis and transportation of the desired radiopharmaceuticals.^{20,32,34,36,37} This relatively long half-life also allows for the radiolabelling of targeting vectors with longer biological half-lives, such as antibodies.³⁴ The drawback of the use of ¹⁸⁶Re for radiotherapy, however, lies in its production. ¹⁸⁶Re is produced by a neutron capture reaction on an enriched source of ¹⁸⁵Re, often resulting in the unavoidable contamination of the desired radionuclide with non-radioactive ¹⁸⁵Re, resulting in non-carrier-free ¹⁸⁶Re.³² The ¹⁸⁶Re isotope can therefore only be produced with low specific activity, however, research is currently being carried out on the cyclotron production of high specific activity ¹⁸⁶Re isotopes.^38–40 The use of ¹⁸⁶Re has been proposed in several applications over the years,^41–44 however, the arguably more popular rhenium radioisotope is ¹⁸⁸Re. While both radioisotopes are effective in irradiating cancerous tissue and have similar chemical properties, ¹⁸⁸Re is often the radiorhenium isotope of choice for two main reasons. The first is that, unlike the production of ¹⁸⁶Re, ¹⁸⁸Re is easily produced by the radioactive decay of ¹⁸⁸W in a ¹⁸⁸W/¹⁸⁸Re generator.^{32,34,36,45–47} The ¹⁸⁸W likely decays to a mixture of mostly ¹⁸⁸Re and a small amount of its metastable radioisotope, ^188mRe, which itself decays to ¹⁸⁸Re with a very short half-life of 18.6 minutes. The ¹⁸⁸W/¹⁸⁸Re generator system is based on the same principle as the ⁹⁹Mo/^99mTc generator from which ^99mTc is eluted. ¹⁸⁸W itself, however, can only be produced by double neutron capture of ¹⁸⁶W in high flux nuclear reactors such as HFIR at Oak Ridge in the USA, the SM3 reactor in Dimitrovgrad, Russia and BR2 in Mol, Belgium.⁴⁵ Although this limits the supply of ¹⁸⁸W for ¹⁸⁸Re production, the generator method of ¹⁸⁸Re production remains particularly attractive, especially since it can produce non-carrier added (nca) radiorhenium isotopes with high specific activities just by a simple elution process, providing daily availability of ¹⁸⁸Re.^34,45,47 The second reason for the popularity of ¹⁸⁸Re is its relatively short half-life of about 17 hours.^3,32,34 Although this restricts its use to biological targeting agents with a similarly short half-life and rapid target uptake,⁴⁸ its relatively short half-life is more appropriate for therapeutic purposes as it poses a lower overall toxicity risk than ¹⁸⁶Re. With the increasing availability of commercial generators, the respective ¹⁸⁸Re radiopharmaceuticals may also conveniently be prepared directly in hospitals, even in remote locations far from nuclear reactors.^46,49 These ¹⁸⁸W/¹⁸⁸Re generators may be able to provide radiotherapeutic treatments to hundreds of patients over their 2–6 month lifetime by ensuring a constant supply of ¹⁸⁸Re.⁴⁹

The use of radiorhenium in the field of nuclear therapeutics has fluctuated over the past three decades, often shadowing the popularity of the radionuclide, ^99mTc. This metastable technetium isotope has been the workhorse of nuclear medicine for decades and is used in approximately 90% of all nuclear imaging applications.^20,29,50,51Fig. 2a provides a graphic image of how extensively studied and reported ^99mTc has been over the years in comparison to the lesser emphasised radiorhenium isotopes, while Fig. 2b tracks the wavering publications of radiorhenium since 1965. Although these radioisotopes could not be more different in terms of the attention they have received, technetium and rhenium have many chemical and structural similarities with regards to complex formation as they are group 7 congeners.^25,27,36,50 Because of these similarities, and the fact that there exists no non-radioactive isotope of technetium, “cold” rhenium has often been used as a replacement for ^99mTc for nonradioactive characterisation and mechanistic elucidation purposes.^29,34,50 Similarly, ^99mTc radiopharmaceuticals may serve as a model for the preparation of novel ^186/188Re radiotherapeutics, allowing for the formation of isostructural complexes of the different isotopes.^29,34,50 The “matched-pair theory,” in which the Tc and Re isotopes are utilised for tandem therapeutic and diagnostic (theranostic) applications, has also become an attractive concept in the field of nuclear medicine.^52–54 This offers the possibility of developing ^99mTc agents for patient pre-selection and follow-up by diagnostic imaging and, thereafter, treatment with the corresponding ^186/188Re analogues.

Fig. 2. Total publications per year on a) ^99mTc, ¹⁸⁶Re and ¹⁸⁸Re and b) ¹⁸⁶Re and ¹⁸⁸Re only (data from the Web of Science, 4 March 2021).

Fig. 2a and b show that, similar to ^99mTc, the development of ^186/188Re chemistry reached a peak around the year 2000, however, publications concerning the use of radiorhenium in clinical research thereafter plummeted. This was largely due to the global shortage at that time of ⁹⁹Mo, the parent radionuclide for the production of ^99mTc, leading to a concomitant crisis in the field of nuclear chemistry.^25,34 Focus was therefore shifted to the use of more convenient chemical isotopes, such as the β-emitting radionuclides ⁹⁰Y and ¹⁷⁷Lu, both of which can be developed at high capacities and do not require elaborate chelator chemistry.^25,34 Although this took the focus off of ¹⁸⁶Re and ¹⁸⁸Re for some time, there has been a resurgence in interest in these radioisotopes in the past decade, as was predicted by Blower in 2017.²⁵ While this deceleration of the downward trend observed in the popularity of radiorhenium is a good start, the acceptance for rhenium isotopes is nowhere near where it once was, nor where it could be. This is largely due to the current availability of the parent radionuclide, ¹⁸⁸W, which is only produced with high enough specific activity in a limited number of high-flux reactors and only on a significant scale at Oak Ridge National Laboratory in the USA. It is nonetheless reassuring to see a number of research groups actively working on radiorhenium-labelled compounds across the globe with the intention of proving their value as radiopharmaceuticals in the field of nuclear medicine.

The design of rhenium radiopharmaceuticals

The design of suitable rhenium-based radiopharmaceuticals is not often a straightforward process. Since rhenium radiopharmaceuticals are developed to target specific organs, any small changes in their ligand structure could drastically affect the in vivo behaviour of the resulting complex and cause the undesirable accumulation of radioactive compounds in the body.¹⁸ A lot of thought therefore needs to be put towards designing radiopharmaceuticals with the desired internal biodistribution, often involving the fine-tuning of their physico-chemical properties.^18,20 Taking this into account, as well as their level of synthetic complexity, rhenium radiopharmaceuticals may be divided into three generations (Fig. 3).^18,20

Fig. 3. Depiction and examples of each of the three generations of rhenium radiopharmaceuticals: a) first generation, b) second generation and c) third generation (BM = biomolecule).

First generation radiopharmaceuticals (Fig. 3a) do not display any targeting capabilities and are composed of ligands systems which are incapable of independently exerting biological functions without the radiometal centre.^18,20,49 Also known as perfusion agents, this generation of radiopharmaceuticals depends entirely on blood flow for biodistribution.^18,20,49 While first generation technetium-99m complexes dominate the field of commercially-available medical imaging agents,¹⁸ several rhenium radiopharmaceuticals of this type have also proved effective in various applications. Such examples include ^186/188Re-HEDP complexes^55,56 and the [¹⁸⁸Re]ReO(DMSA)₂ (DMSA = dianionic dimercaptosuccinic acid) complex⁵⁷ (Fig. 3a), which were all developed for the palliative treatment of painful bone metastases. As a result of their limited targeting capabilities and the “trial and error” approach in their conceptualisation, focus has generally shifted away from first generation radiopharmaceuticals to more sophisticated and specific targeting agents.

Second and third generation radiopharmaceuticals (Fig. 3b and c) are both capable of targeting and binding to a specific receptor.^18,20,49 The former makes use of a biological molecule, such as a peptide, protein or antibody, as the targeting vector to which the radiometal is attached. This is achieved through the use of the previously-discussed BFC, which couples the biovector to the radiometal chelate via a variable linker.^18,49,58 A well-known example of a BFC used in both rhenium and technetium radiochemistry is MAG₃ (Fig. 3b), which has a pendant carboxylate for biovector conjugation.^18,20 The square pyramidal ^186/188Re(v) complex of this ligand is used most commonly as the radiation source during balloon angioplasty for the prevention of restenosis, as it has the favourable property of fast renal excretion.^20,59,60 Third generation radiopharmaceuticals, on the other hand, incorporate the radiometal into the carbon skeleton of the ligand in order to mimic the essential structures of biomolecules, such as the steroid hormones, testosterone, oestrogen and progesterone.^18,20,58Fig. 3c shows the incorporation of the oxorhenium(v) complex into the structure of progesterone, replacing the B and C rings as shown.^58,61 Although bidentate chelators such as those shown in Fig. 3c often do not stabilise the Re-centre enough in vivo, these mimicked biomolecules are meant to bind themselves to an intrinsic receptor without the need for attachment to a separate receptor-binding biovector, as in the case of second generation radiopharmaceuticals. Since it is an extremely challenging task to imitate exactly the various physical and chemical properties of hormones using metal complexes, the current focus in the therapeutic medicine domain is mainly on the development of suitable bifunctional chelators for second generation radiopharmaceuticals. This research has been very active over the years in the context of finding ideal chelating systems for radiorhenium, although no radiorhenium complexes for targeted radiopharmaceutical therapy have been introduced to the market for commercial use. Many factors such as the attachment site of the biomolecule, the length of the linker and the overall size, charge and solubility of the complex may affect the efficacy of the radiopharmaceutical as a whole.⁴⁹ However, once all these factors can be optimised, there is great potential for the use of the BFC approach in the design of novel rhenium targeted radiopharmaceuticals.

Past and present approaches in the design of bifunctional chelators (BFCs) for radiorhenium

Coordination complexes of rhenium are extremely diverse, owing to the ability of rhenium to exist in a wide range of oxidation states (−1 to +7). The higher oxidation states require the coordination of π-donating ligands, namely oxido, imido and nitride, in order to counteract the electron-deficient metal centre.^32,62 Conversely, good π-acceptor ligands like carbonyls or tertiary phosphines are required to stabilise rhenium in its lower oxidation states.^32,62 The synthesis of all rhenium radiopharmaceuticals begins with an aqueous solution of rhenium in its +7 state as the perrhenate anion, ReO₄⁻.^16,32 ReO₄⁻ is more difficult to reduce than its analogous pertechnetate anion, TcO₄⁻, evidenced by the significant difference in their respective standard reduction potentials (ReO₄⁻/ReO₂ = 0.510 V and TcO₄⁻/TcO₂ = 0.738 V).^63,64 This difference in reduction potentials means the methods required to develop novel radiorhenium therapeutic agents cannot always follow the same routes applied for the synthesis of technetium complexes.^32,62 Often harsher conditions are required, which is one of the main stumbling blocks in the development of novel rhenium radiopharmaceuticals.⁴⁵ BFCs developed for radiorhenium must therefore take the kinetic, thermodynamic and redox stabilities of the final complex into account, ensuring that this complex is able to display high chemical inertness and stability, particularly under physiological conditions.¹⁶ Additionally, the choice of the most stable metal core is also vital in forming strong rhenium chelates. According to the oxidation state of this rhenium core, a brief summary of some of the past chelator successes and failings will be discussed, as well as current advances in the design of suitable BFCs developed for ^186/188Re.

Bifunctional chelators for ^186/188Re(v)

Similar to Tc(v) chemistry, the most extensively studied BFCs developed for Re(v) incorporate the oxorhenium(v) [ReO]³⁺ core.¹⁶ These chelators are generally of the N_xS_4−x tetradentate donor type. Chelators of this type are able to stabilise the oxorhenium(v) core through the strong π-bonding oxo-group and σ-bonding thiolate, amino and amido groups of the chelator moiety, with the latter also providing some π-bonding contributions.^14,16,65 As a result, these oxorhenium complexes adopt a square-pyramidal geometry with the tetradentate chelating ligand occupying the equatorial plane and the oxo-group in the axial position.⁶⁶ Alternatives to the oxorhenium(v) core are the lesser-studied dioxo [ReO₂]⁺ and nitride [ReN]²⁺ cores, around which various BFCs have also been designed.²⁵ A common setback in the design of radiorhenium(v) radiopharmaceuticals is their susceptibility to oxidation to perrhenate in vivo.^16,65 This is a persistent problem and remains largely the sole reason why more Re(v) complexes have not been taken forward for further testing as radiopharmaceuticals. The BFCs for Re(v) must therefore be carefully selected and thoroughly tested to ensure optimal in vivo stability.

N₃S chelators

The triamidomonothiol (N₃S) chelator framework has been a bifunctional chelating system widely used for ^99mTc, but has also shown popularity with ^186/188Re.^{23,59,67–70} Since many antibodies and proteins are heat-sensitive, the pre-conjugation labelling approach (i.e. radiolabelling of the BFC prior to biomolecule conjugation) is often adopted.²³ Following this strategy, the first N₃S chelator used for radiorhenium was based on the ^99mTc(v)-labelled free mercaptoacetyltriglycine (MAG₃) complex, which showed success in clinical studies for the diagnosis of renal and urinary disorders.²³ Although the ^99mTc-MAG₃ complex is stable, the MAG₃ chelator, used as is, produced very low complex yields when coordinated to radiorhenium.⁷¹ As a result, in 1989 the MAG₂-GABA (MAG₂-γ-aminobutyrate) chelator, modified from MAG₃, became the first N₃S BFC for radiorhenium preconjugation labelling of antibodies (Fig. 4a).^71,72 This was later replaced with the S-benzoyl-MAG₃ chelator (Fig. 4b), which was first radiolabelled with ¹⁸⁶Re in 1993 and continues to be used today as an effective way of producing ^186/188Re-MAG₃.^73–75 The extensive use of this version of the MAG₃ chelator in nuclear medicine can be attributed to its ease of synthesis, long shelf-life and stable rhenium complex produced, particularly in vivo.⁷⁶ More recently, in 2017 Castillo Gomez et al. reported the synthesis of N₃S-tetradentate thiocarbamoylbenzamidine BFCs to establish technetium and rhenium complexes with similar properties to MAG₃ (Fig. 5).⁷⁷ The advantage of their ligand over MAG₃ was the presence of the thiocarbonyl sulphur atom instead of a thiol group, which would require protection and deprotection.⁷⁷ Although only non-radioactive rhenium complexes conjugated to the peptide hormone, angiotensin-II, were synthesised, the good in vivo stability of these compounds makes them candidates for future testing as radiopharmaceuticals with radiorhenium.

Fig. 4. N₃S-Based BFCs, (a) MAG₂-GABA and (b) S-benzoyl-MAG₃, used with the [^186/188Re O]³⁺ core.

Fig. 5. General structure of the N₃S-tetradentate thiocarbamoylbenzamidine BFCs used to establish rhenium complexes by Castillo Gomez et al.⁷⁷.

A disadvantage of pre-conjugation labelling is the extensive post-labelling purification required for the radioactive radiopharmaceuticals.^23,69 Post-conjugation labelling is therefore the simpler and more practical approach when the biomolecule in question can withstand harsh environments. There are a number of examples of post-conjugation radiolabelling of N₃S BFCs in literature, with some stable peptides interestingly able to act as both the chelating moiety and the targeting vector within the same molecule.^23,78–81 A recent example of this is the [¹⁸⁸Re]ReO-KYCAR complex made by Sanders et al. in 2019.⁷⁸ They prepared various Tc and Re complexes of the KYCAR (lysyl-tyrosyl-cystyl-alanyl-arginine) pentapeptide sequence (Fig. 6), with chelation to the radiometal centre made possible through the intrinsic N₃S donor atoms.⁷⁸ While all the radiometal complexes were synthesised with high purity of greater than 95%, the [¹⁸⁸Re]ReO-KYCAR complex showed 50% decomposition in vitro in both PBS and human serum.⁷⁸ This instability was also evident in vivo and was determined to be the result of oxidation to ¹⁸⁸ReO₄⁻, a common trend observed with radiorhenium complexes.

Fig. 6. Structure of the KYCAR N₃S-pentapeptide ligand used to establish ^99mTc, ¹⁸⁸Re and ^natRe complexes by Sanders et al.⁷⁸.

While research into the literature shows that N₃S systems can be used to chelate oxorhenium(v), these complexes still do not have the same excellent stability as oxotechnetium(v) complexes on the radiotracer level, likely due to oxidation of the radiometal.^38,82 The addition of a second thiolate group to form an N₂S₂-type chelator may therefore provide this stability, particularly in vivo.

N₂S₂ chelators

While many of the BFCs for radiorhenium are based on the N₃S ligand framework, the N₂S₂ chelating system has also garnered considerable interest since it is able to form stable complexes with radiorhenium(v).⁸³ This class of ligands includes bis(aminoethanethiols) (BAT),⁸⁴ monoamino-monoamidodithiols (MAMA)^85,86 and diamidodithiols (DADS),^83,87 some rhenium complex examples of which are shown in Fig. 7. Of these, BAT and MAMA ligands are able to form neutral and highly lipophilic complexes with [Re O]³⁺ and are used in various applications where hydrophobic targeting is required. While the more lipophilic BAT chelator has been used in imaging agents which target various brain receptors, MAMA chelators have most commonly been developed for tumour therapeutics.¹⁶ In terms of the use of N₂S₂ ligands as BFCs, MAMA chelators have more often been favoured because of their single amine functionality for easy bioconjugation.¹⁶ In fact, MAMA-type BFCs of radiorhenium have been conjugated to many types of targeting vectors with great success, from antibodies to peptides and small molecules (Fig. 8).^{14,34,88–93} The ¹⁸⁸Re-labelled tetrapeptide, [¹⁸⁸Re]Re–N₂S₂-IMP-192 synthesised in 2000, has been used for tumour pretargeting and showed promising results in colorectal cancer radioimmunotherapy (Fig. 8a).^14,88 Most importantly, biodistribution studies of this ¹⁸⁸Re-radiolabelled peptide showed a tumour/non-tumour ratio of between 3.14 and 22.5 to 1 for every major organ.⁸⁸ Similarly high tumour-uptake ratios were achieved in 2009 with the ¹⁸⁸Re-labelled antibody, trastuzumab, via the N₂S₂ BFC, succinimidyl 3,6-diaza-5-oxo-3-[2-((triphenylmethyl)thio)ethyl]-8-[(triphenylmethyl)thio]octanoate (SOCTA)⁸⁹ (Fig. 8b). This ¹⁸⁸Re-SOCTA–trastuzumab complex showed therapeutic potential as a radiopharmaceutical in the treatment of breast cancer.⁸⁹ In the treatment of painful bone metastases, the ¹⁸⁶Re-MAMA-HBP complex (Fig. 8c) synthesised by Ogawa et al. displayed a greater affinity for bones than the original ¹⁸⁶Re-HEDP radiopharmaceutical, which did not make use of a BFC.^90,94 Finally, Tang et al. in 2011 also found success with a novel ¹⁸⁸Re-labelled N₂S₂ bifunctional chelator, H3MN-16ET, for the treatment of hepatocellular carcinoma (Fig. 8d).^91,92 Upon treating rats that had hepatic tumours with the ¹⁸⁸Re-MN-16ET complex conjugated to lipiodol (poppy oil lipid used for embolization of the tumour's blood vessels), a decreased tumour size and increased rat survival rate compared to lipiodol alone were observed.^91–93 In all of these examples, the N₂S₂ chelator plays a vital role in securing the radioactive rhenium to the targeting vector, thereby successfully preventing the loss of the radionuclide in vivo.

Fig. 7. Representative examples of N₂S₂ chelating systems used with the oxorhenium(v) core.

Fig. 8. Examples of MAMA-type N₂S₂ BFCs of radiorhenium that have been conjugated to various targeting vectors.

More recently, Demoin et al. in 2016 synthesised novel 222-MAMA-based BFCs that were coordinated to ^186/188Re to be tested as potential radiopharmaceuticals.³⁵ The 222-MAMA chelator was chosen because of the study by Oya et al. in 1998, which found 222-MAMA ligands chelated to ^99mTc(v) to be more stable over time than their N₂S₂-BAT ligand counterparts.⁹⁵ This is likely due to the greater hydrophilicity of the MAMA-based complexes. Computational studies performed by Demoin et al. were able to confirm that these MAMA complexes are more thermodynamically stable than their BAT analogues.⁹⁶ In the research done by Demoin, a bombesin peptide, BBN(7–14)NH₂, was conjugated to the 222-MAMA ligand via an Ahx linker (Fig. 9a) and thereafter, radiolabelled with ^99mTc and ¹⁸⁶Re (carrier-added and no-carrier-added).^35,96 The [¹⁸⁶Re]ReO-222-MAMA(N-6-Ahx-BBN(7–14)NH₂) complex was found to bind with high affinity to the GRP receptors that are overexpressed on the surface of PC-3 human prostate cancer cells.^35,96 Unfortunately, its high lipophilicity caused this complex to be excreted via the hepatobiliary and GI system over the renal system and, very recently in 2020, Khosroshahi has focused on synthesising a 222-MAMA BFC for ¹⁸⁶Re with improved hydrophilicity (Fig. 9b).⁹⁷

Fig. 9. 222-MAMA BFCs conjugated to bombesin synthesised by: a) Demoin^35,96 and b) Khosroshahi.⁹⁷.

N₄ and polyaminopolycarboxylate chelators

While tetraamine (N₄) chelators have been used extensively, both in their cyclic and acyclic forms, to attach ^99mTc to biomolecules, they have only very rarely been used for ^186/188Re radiolabelling.^22,23 Rhenium complexes of the N₄ chelating system are six-coordinate, incorporating a cationic dioxorhenium(v) core as depicted in Fig. 10.⁹⁸ Since these N₄ chelating moieties are largely lipophilic, making use of the [O Re(v) O]⁺ core could improve the overall hydrophilicity of the complex.⁹⁹ This would be useful in accelerating the excretion of the radiopharmaceutical via the kidneys and urinary system.⁹⁹ Some early examples of N₄ chelator complexes of radiorhenium are those studied in 1996 by Prakash et al., who assessed the ease of synthesis and kinetic stability of several tetradentate N₄-based ligands.⁹⁸ Two out of the six ligands studied, one cyclic (cyclam) and one acyclic (Fig. 10), showed no degradation in human serum 24 hours after their complexation to rhenium, attesting to their excellent stability.⁹⁸ Prakash concluded that by fine-tuning their synthesis and biological stability, these dioxorhenium(v) complexes of the N₄ tetradentate ligand system have great potential for use as radiopharmaceuticals.

Fig. 10. Two N₄-chelators studied by Prakash et al., one cyclic and one acyclic, coordinated to the dioxorhenium(v) core.⁹⁸.

Since their publication, other types of N₄ chelators for rhenium(v) have been attempted, such as the pyrazolyl-based N₄ chelator synthesised by Moura et al. in 2006 (Fig. 11), but have had minimal success with radiorhenium.¹⁰⁰ The complex formed by the trans-[ReO(OMe)]²⁺ core coordinated to the pyrazolyl-triamine ligand was unfortunately determined to have no relevance in the field of radiotherapeutics.¹⁰⁰ This is because this unique monooxorhenium(v) complex is not achievable under the aqueous conditions required to synthesise radiopharmaceuticals and the dioxorhenium(v) complex of this ligand was unable to be isolated.¹⁰⁰ This result shows why the use of N₄-based BFCs to label biomolecules with radiorhenium is rare in more recent times.

Fig. 11. Structure of the mono-oxorhenium(v) complex with the trans-[ReO(OMe)]²⁺ core synthesised by Moura et al.¹⁰⁰.

A similar issue of reoxidation of the rhenium core has also been observed with polyaminopolycarboxylate ligands – a chelating system which is well known to radiochemists for coordinating radioisotopes of copper, indium and yttrium. While ¹⁸⁸Re-labelled free DTPA (diethylenetriamine pentaacetate) can easily be produced and has shown success in endovascular brachytherapy,^101,102 the attempts made with radiolabelling a DTPA-conjugated antibody with ¹⁸⁶Re were unsuccessful.^23,103 The radiolabelling efficiency was very low, at less than 18%, and the ¹⁸⁶Re-DTPA-antibody complex showed instability in vivo.^23,103 Additionally, the structure of the resulting radiorhenium-DTPA complexes remain unknown. Radiochemists have thus moved away from both tetraamine (N₄) and polyaminopolycarboxylate chelating systems for radiorhenium and have made use of other coordinating atoms, such as phosphorous, sulphur and oxygen, in an attempt to better stabilise the Re(v) core.

Phosphorous-, sulphur- and oxygen-donor atoms and the Re(v)-nitrido core

Athough ligands of the type N_xS_4−x dominate the field of BFCs for ^186/188Re, many issues have emerged as a result of their physicochemical properties and ease-of-radiolabelling. The high lipophilicity of these chelators affects the receptor-binding and tissue clearance capabilities of the radiolabelled biomolecule, while harsh conditions are often required in the radiolabelling process. To overcome these obstacles, Volkert et al. in 1999 developed the first water-soluble, BFC based on a P₂S₂–COOH framework (Fig. 12a), since previous studies had shown that phosphine-containing chelators form strong and stable complexes with Tc and Re in high radiochemical purity.^104,105 The excellent hydrophilicity of the P₂S₂-chelate can be attributed to the overall positive charge of the compound, as well as the presence of the hydroxymethyl groups attached to the phosphine moieties.^104,105 Upon conjugation of this BFC to a peptide, followed by postconjugation labelling with ¹⁸⁸Re, the resulting radiocomplex exhibited good stability in vitro and in vivo with very minimal in vivo oxidation to the perrhenate anion.^104,105 The hydrophobic nature of the two thioether functionalities, however, prompted their replacement with amide groups in a subsequent study in 2002, forming a diamido-dihydroxymethylene phosphine (N₂P₂–COOH) bifunctional chelator (Fig. 12b).¹⁰⁶ As expected, both the ^99mTc and ¹⁸⁸Re complexes of this chelator exhibited efficient bloodstream clearance via the kidneys, instead of through the hepatobiliary system as experienced with the P₂S₂-type BFCs.¹⁰⁶ These excretion results in particular make the N₂P₂-type BFC attractive for the radiolabelling of peptides in future studies.

Fig. 12. Structuresof the water-soluble ¹⁸⁸Re complexes by Volkert et al. which incorporate BFCs of the type (a) P₂S₂ (ref. 104) and (b) N₂P₂.¹⁰⁶.

More recently, an alternative approach called the “metal fragment” strategy has been a way to prepare BFCs for radiorhenium which are inert to oxidation to ReO₄⁻.¹⁰⁷ Following this strategy, a “robust” precursor metal complex is prepared having specific ligands which stabilise the metal centre and are inert towards oxidation–reduction reactions.¹⁰⁷ Additional substitution-labile ligands are also incorporated which can be easily replaced by ligands that have stronger coordination systems and which are attached to biomolecules, resulting in a stable, bioactive radiopharmaceutical.¹⁰⁷ A number of rhenium mixed-ligand complexes have been prepared as potential targeted radiopharmaceuticals, many of which incorporate the nitride core, [Re N]²⁺. There is a multistep radiolabelling approach to synthesising compounds of this type, with the [^186/188Re]Re(v)-nitride core being generated in the first step by the reaction of ^186/188ReO₄⁻ with a hydrazide derivative and in the presence of a reducing agent, such as SnCl₂·2H₂O.^25,107,108 This intermediate ^186/188Re(v)-nitride complex is then treated with the suitable ligands which stably coordinate to the metal core, producing a final, square-pyramidal complex.^25,107,108 Although a number of hydrazide derivatives have been evaluated as the N³⁻ source in the past,¹⁰⁹ the most preferred source is N-methyl-S-methyl dithiocarbazate (HDTCZ) as shown in the two-step process in Fig. 13.^25,107–109Fig. 13 shows the synthesis of the bis(diethyldithiocarbamato) nitride ¹⁸⁸Re(v) complex ([¹⁸⁸Re]ReN-DEDC), a highly lipophilic complex which was used to label lipiodol for the treatment of hepatocellular carcinoma and displayed excellent preferential uptake in the lesion.¹⁰⁷

Fig. 13. Two-step procedure for the synthesis of the ¹⁸⁸Re(v)-nitrido complex, [¹⁸⁸Re]ReN-DEDC.¹⁰⁷.

On the basis of the “mixed ligand” strategy, some of the most commonly utilised Re(v)-nitride systems include the [Re(N)(PNP)]²⁺ (PNP = phosphinoamine ligand) system^109,110 and the “3 + 1” mixed-ligand system, with the latter being popular because of the great versatility in its approach.¹⁰⁷ Surrounding the apical [Re N]²⁺ core, the essential features of the “3 + 1” complexes include an anionic, tridentate ligand (XYZ) and a monodentate tertiary phosphine ligand (PR₃), as depicted in Fig. 14a.¹⁰⁷ The tridentate XYZ ligand should be a good π-donor and have three donor atoms of the type S or N, while the monophosphine ligand should be a good π-acceptor.¹⁰⁷ This is exemplified by the ¹⁸⁸Re peptide conjugate prepared by Smilkov et al. in 2014, the structure of which is shown in Fig. 14b.¹¹¹ This BFC was synthesised following the “3 + 1” strategy to label substance-P (SP), an undecapeptide member of the tachykinin neuropeptide family that is a major endogenous ligand for the NK1 receptor type.^107,111 Since these NK1 receptors have been found on the surfaces of several types of cancerous tumours, such as malignant brain tumours and pancreatic carcinoma among others, targeted radionuclide therapy could be used as a way of targeting various tumour-specific SP-NK1 receptors.^107,111 The [¹⁸⁸Re]Re-SP radioconjugate synthesised by Smilkov et al. displayed an excellent affinity for the cell-surface NK1 receptors on U-87 MG human glioblastoma cells upon incubation with this cell-line.^107,111 A similar group of “3 + 1” ¹⁸⁸Re chelator complexes encompassing the [Re N]²⁺ core was developed for the attachment to biotin for use in intra-operative avidination for radionuclide therapy (IART).¹¹² IART is an approach used for accelerated radiotherapy in breast cancer patients and the resulting ¹⁸⁸Re-biotin conjugates displayed high in vitro stability and in vivo inertness, particularly towards biotin-degrading enzymes.¹¹² In fact, this therapeutic chelator was so successful that a lyophilised ready-to-use kit was developed for the easy, on-site preparation of the radiocompound in hospitals.¹¹²

Fig. 14. (a) General structure of the nitride ¹⁸⁸Re “3 + 1”complex and (b) an example of the ¹⁸⁸Re-nitride “3 + 1”strategy to label substance-P.¹¹¹.

More recently, various novel Re(v) complexes have been synthesised following the “mixed ligand” strategy with the intent of synthesising their radiorhenium congeners for radiotherapy, although very few of these radioactive compounds have come to fruition.^113,114 While the [Re N]²⁺ core is stable and very versatile, novel radiorhenium therapeutic agents of this type has not found widespread use, likely due to the many steps required in the labelling process of these compounds.

HYNIC and its derivatives

Today, one of the important alternatives to the N_xS_4−x system of BFCs is the HYNIC (6-hydrazinonicotinic acid) ligand system, the structure of which is shown in Fig. 15a.¹⁰⁴ First synthesised by Abrams et al. in 1990 for the ^99mTc-labelling of human polyclonal IgG, this popular ligand has since been adopted for numerous radiolabelling applications by the manipulation of the carboxylate group for biomolecule attachment.^66,115,116 Strictly speaking, HYNIC is not a chelator as its strong association to the metal centre is through the formation of a single diazenido bond with the metal and not via the chelation effect.¹¹⁷ This strong metal-binding to form a robust core does however mean that HYNIC-attached biomolecules will not release the radionuclide in vivo.^22,66,118 Since it is only able to coordinate to a metal through a maximum of two donor groups (the hydrazine and pyridyl nitrogens), HYNIC requires a set of co-ligands to fully saturate the coordination sphere of the metal.¹¹⁸ By varying the co-ligands used to stabilise the metal oxidation state, the in vivo metabolism and excretion properties as well as the pharmacokinetic properties of the radiocompound can be modified. Co-ligands such as tricine, water-soluble phosphines, nicotinic acid and EDDA are often chosen, however, a limitation to their use lies in the formation of stereoisomers upon metal coordination, each of which may have different biodistribution properties. Additionally, the mode of coordination of HYNIC to the metal centre is not known. It may bind in a monodentate fashion through the terminal hydrazine nitrogen (Fig. 15b), but most likely forms bidentate complexes via an additional pyridyl nitrogen coordination (Fig. 15c). This was shown to be the case for many of the crystallographically-characterised complexes of Tc and Re with HYNIC-type ligands such as 2-hydrazinopyridine.^22,118,119 It is this vagueness in the overall structure of the HYNIC-metal complex which often detracts from the attractiveness of this ligand for use in a clinical setting.

Fig. 15. (a) Structure of HYNIC and its binding modes, (b) monodentate or (c) bidentate.

Rhenium complexes of HYNIC have been extensively investigated, since non-radioactive rhenium is often used in place of ^99mTc, which itself lacks a stable isotope.²² In spite of this wealth of knowledge pertaining to radiolabelling biomolecules with HYNIC as a BFC, there are very few reports published on the use of radiorhenium compared to ^99mTc and ^natRe. In fact, [¹⁸⁸Re]Re-HYNIC-4B4 (ref. 120) in 2006 and, more recently, [¹⁸⁸Re]Re-HYNIC-trastuzumab¹²¹ and [¹⁸⁸Re]Re-HYNIC-SP94 (ref. 122) in 2015 and 2016, respectively, appear to be the only examples of radiorhenium-peptide complexes incorporating HYNIC in literature. While these rhenium-radiolabelled peptides displayed good specific binding to their desired targets and were deemed as potential therapeutic radiopharmaceuticals, many other attempts at radiolabelling biomolecules with radiorhenium using HYNIC were unsuccessful. This is largely due to the susceptibility of radiorhenium to air oxidation, with the labelling efficiency of a previously-reported [¹⁸⁸Re]Re-HYNIC conjugate dropping from 97% to 80% over 1 hour in storage as a result of air oxidation.¹²³ Additionally, the ambiguity that arises as a result of the various coordination modes of HYNIC, as well as the formation of mixtures of isomers because of the co-ligands used, have also made isolating pure [¹⁸⁸Re]Re-HYNIC-peptide conjugates difficult. For these reasons, the HYNIC ligand was subsequently modified at the hydrazinic nitrogen by extension with a thioamide tethering group to produce a ligand system capable of forming more stable and well-defined Re(v) complexes.

In a preliminary communication published in 2004, Clarke et al. reported the synthesis of pyridylphenylthiocarbazide, a novel thioamide derivative of pyridylhydrazine which is known as “SHYNIC” for short as it is a sulphur-containing derivative of HYNIC (Fig. 16a).¹²⁴ Two of these SHYNIC ligands were coordinated to a Re(v) oxo core in a bidentate NS fashion, forming a stable, well-defined rhenium complex in a high yield (Fig. 16b).^16,124 Additionally, this result showed there was no longer a need for co-ligands to synthesise complexes of this type, thereby reducing the undesirable occurrence of geometrical isomers.¹⁶ This concept was extended by North et al. in 2017 whereby a family of substituted SHYNIC ligands conjugated to various peptides was used to prepare complexes with the [M O]³⁺ core (where M = ^99mTc, ^natRe or ¹⁸⁸Re), as shown in Fig. 17.⁵⁰ A full set of ^99mTc and ^natRe complexes were prepared with two different types of targeting peptides – cyclic arginine-glycine-aspartic acid fibronectin fragment (cRGDfK) and Tyr³-octreotate (Tyr³-Oct).⁵⁰ Only one ¹⁸⁸Re complex was reported with Tyr³-Oct as the targeting peptide and was produced with a ∼67% yield from [¹⁸⁸ReO₄]⁻, although this yield could be improved via reaction condition optimisation.⁵⁰ It was concluded that these new SHYNIC-containing complexes warrant further investigation as potential theranostic agents, utilising a ^99mTc imaging and ^188/186Re therapeutic matched pair.⁵⁰

Fig. 16. Structures of (a) SHYNIC and (b) a Re(v) oxo complex featuring two coordinated SHYNIC ligands.¹²⁴.

Fig. 17. Synthesis of the SHYNIC complexes of ^99mTc, ^natRe and ¹⁸⁸Re conjugated to cRGDfK and Tyr3-Oct.⁵⁰.

The SHYNIC ligand has only come up again once more in literature since then, when Fletcher et al. in 2018 synthesised six new SHYNIC ligands containing functional groups such as benzofuran or styrylpyridyl.¹²⁵ While these ligands were not conjugated to biomolecules, the functional groups they contained were known to selectively bind amyloid-β plaques, which are deposited in cerebral blood vessels as a result of ageing.¹²⁵ Well-defined Re(v) oxo complexes of these ligands were made as surrogates for ^99mTc, with one of these rhenium complexes (Fig. 18) being shown to bind to amyloid-β plaques within humain brain tissue.¹²⁵ While no radiorhenium complexes were reported in this study, it does add to the pool of knowledge on rhenium complexes of this type.

Fig. 18. Structure of one of the Re(v)-oxo complexes prepared by Fletcher et al. featuring SHYNIC ligands known to selectively bind to amyloid-β plaques.¹²⁵.

Bifunctional chelators for ^186/188Re(iii)

The trivalent state is a very stable and common oxidation state for technetium and rhenium, however, there is a paucity of research into the use of M^III (^99mTc/^186/188Re) complexes as radiopharmaceuticals. This is largely due to the difficulty in preparing these radiocomplexes in aqueous conditions, with the methods used to prepare such compounds at the macroscopic level being ineffective for their preparation at the tracer level.⁵² With respect to nuclear medicine, the chemistry of Tc(iii) and Re(iii) complexes remains largely unexplored, with far fewer examples of radiorhenium(iii) complexes in literature due to the difficulty to reduce Re compared to Tc.¹⁴ For this reason, only a handful of radiorhenium(iii) complexes for radiotherapy have been investigated in contrast to those of ^natRe(iii) (a non-radioactive substitute for ^99mTc), although the latter allows for the potential design of radioactive analogues. Many of the Re(iii) complexes for radiotherapy in literature follow the “molecular fragment” approach, where the complex is made up of substitution-inert ligands as well as labile co-ligands which can be easily exchanged for bifunctional ligands to which biomolecules may be attached. Of these, [Re(PS)₂]⁺ (PS = 1,2-phosphinothiolate) has been shown to be a stable Re(iii) core whose coordination sphere is completed by the inclusion of a bidentate ligand such as dithiocarbamate (Fig. 19a).⁵² Other co-ligands that have been studied with this core include pyridine-2-thiolate and xanthate.¹²⁶ Although only ^natRe(iii) complexes were evaluated in these studies, the ease of synthesis of the corresponding [^99mTc][Tc(PS)₂(L)] complexes, as well as their excellent in vitro and in vivo stability is promising for the design of analogous radiorhenium complexes.^52,126 Additionally, the dithiocarbamate ligand may also act in a bifunctional manner by attaching to a biomolecule as well as the M(iii) core, paving the way towards the potential use of radiorhenium complexes of this type in radiopharmaceutical applications.¹²⁶

Fig. 19. Structures of (a) complexes of the type [M(PS)₂(L)] (M = ^99mTc(iii) and Re(iii))¹²⁶ and (b) the [¹⁸⁸Re]Re(iii)-SSS complex.⁹¹.

Similarly, several examples of “sulphur rich” Re(iii) and Tc(iii) complexes have also followed this multi-ligand approach, such as the [¹⁸⁸Re]Re–(S₂CPh)(S₃CPh)₂ complex (or the [¹⁸⁸Re]Re-SSS complex).¹²⁷ This radiorhenium(iii) core surrounded by six coordinated sulphur atoms from three ligands (Fig. 19b) was used in 2004 to radiolabel lipiodol for the treatment of hepatocellular carcinoma and showed elevated uptake and retention in hepatic tissue when investigated in animal models.¹²⁷ Inspired by the promising biological properties of this complex, Lepareur et al. in 2005 synthesised a related series of M(iii) (M = ^99mTc/^natRe) complexes in order to study their synthetic pathway.¹²⁸ They concluded that, while the two trithioperoxybenzoate ligands are substitution-inert, the labile dithiocarbamate ligand (which replaced the previous dithiocarboxylate ligand) offers the possibility of introducing appended biologically active molecules in future studies.¹²⁸

Another promising approach to radiolabelling biomolecules with radiorhenium is through use of the “4 + 1” mixed-ligand system. In this approach, a tetradentate tris(2-mercaptoethyl)amine (NS₃) ligand and a monodentate PR₃ or isocyanide ligand is coordinated to a Re(iii) core, forming a trigonal bipyramidal complex as shown in Fig. 20.¹²⁹ The monodentate ligand is then able to serve as an attachment site for coupling target-specific biomolecules to the radiorhenium core.¹²⁹ The three thiolate groups of the NS₃ ligand are lipophilic and therefore prevent the hydrolysis and reoxidation of the Re(iii) core to perrhenate.¹²⁹ It was, however, found that the incorporation of a carboxylate group into the tetradentate ligand on the periphery of the complex improved the hydrophilicity of the overall complex without affecting the lipophilic core.^129,130 This resulted in ¹⁸⁸Re(iii) complexes that are very stable at both high and low activity levels in a variety of aqueous media such as phosphate buffer and rat and human plasmas.^129,130 Although this research shows the potential these ¹⁸⁸Re(iii) complexes have in the design of novel radiorhenium therapeutic agents, lately, this NS₃ “4 + 1” approach has been adopted almost exclusively for the development of ^99mTc imaging agents.^131–133

Fig. 20. The “4 + 1”mixed ligand approach.

Most recently, Re(iii) complexes with tetradentate N₂O₂ Schiff base ligands have emerged as analogues of ^99mTc-furifosmin, a successful myocardial perfusion imaging agent (Fig. 21a).^134–137 These types of compounds, known as “Q-complexes,” are made up of a M(iii) metal core to which an equatorial N₂O₂ Schiff base ligand and two trans PR₃ ligands are attached, forming cationic, lipophilic complexes.¹³⁵ Baumeister et al. in 2018 were the first to report Re(iii) analogues of ^99mTc-furifosmin (Fig. 21b) for the potential radiotherapy of multi-drug resistant tumours, as d⁴-type Re(iii) complexes were anticipated to be kinetically inert in biological conditions.¹³⁵ As expected, these water-soluble complexes were stable and resistant to degradation after more than a week.^135,137 A setback in the development of ^186/188Re(iii) Q-complexes, however, is the required use of ^186/188ReO₄⁻, the precursor which often requires much harsher reaction conditions to yield lower radiolabelling yields than ^99mTcO₄⁻.^135,137 Although only ^natRe(iii) Q-complexes have been synthesised to date, the same microwave-assisted synthesis method used for their production may be employed in the future when translating ^99mTc(iii) complexes of this kind to ^186/188Re(iii).^135,137 Baumeister et al. in 2019 also investigated the stability of mixed-ligand Re(iii) and ^99mTc(iii) complexes with bifunctional Schiff base ligands by functionalising the Schiff base backbone.¹³⁶ The Schiff base complexes incorporating benzamide-linked biomolecules displayed superior stability in solution and were therefore proposed as a potential class of bifunctional radiopharmaceuticals in the future.¹³⁶

Fig. 21. Structures of (a) ^99mTc-furifosmin and (b) the Re(iii) analogue synthesised by Baumeister et al.¹³⁵.

Bifunctional chelators for ^186/188Re(i)

While complexes containing the Re(v) core remain the most commonly exploited in the design of [^186/188Re]Re-radiopharmaceuticals, a relatively new approach makes use of a Re(i) core. The majority of research done on complexes of this oxidation state comprises the metal-tricarbonyl system, with predominantly “cold” rhenium-tricarbonyl complexes being synthesised as model complexes for structural analysis of the analogous [^99mTc]Tc-complexes. BFCs have then been designed to be incorporated into this core to take up the remaining coordination sites not already occupied by carbonyls. The remainder of the Re(i) complexes in literature which do not contain the tricarbonyl ligand system include mostly rhenium bis-arene complexes, which can be functionalised on the arene rings for bioconjugation to targeting molecules.^138,139 No radiorhenium bis-arene complexes have been synthesised to date, however, the stable and well-protected metal core makes these sandwich complexes ideal for future radiotherapy research. The [¹⁸⁶Re]Re(i)-analogue of the myocardial perfusion imaging agent, [^99mTc]Tc-MIBI (Cardiolite®), has also been reported and displays another interesting core for biomolecule functionalisation in the future (Fig. 22).¹⁴⁰

Fig. 22. The structure of [¹⁸⁶Re]Re(i)-MIBI (MIBI = 2-methoxyisobutyl-isonitrile).

The [Re(CO)₃]⁺ core

Since the first reporting of the [^99mTc][Tc(OH₂)₃(CO)₃]⁺ moiety by Alberto and co-workers in 1998, both Tc(i) and Re(i) tricarbonyl complexes have been studied extensively.^31,141–143 The favourable properties of the fac-[M(CO)₃]⁺ (M = ^99mTc(i), ^186/188/natRe(i)) core make it a very versatile synthon for radiolabelling biomolecules, with the low-spin d⁶ electronic configuration rendering the metal centre kinetically inert and therefore very stable for medical applications.^16,36 The precursor to the carbonyl core exists in the form of [M(OH₂)₃(CO)₃]⁺ (where M = T(i) or Re(i)). The ^99mTc analogue of the precursor can be synthesised directly from [^99mTcO₄]⁻ in saline using a boron-based carbonylating agent, K₂[H₃BCO₂], as the CO source and can be obtained from a commercially available kit (IsoLink™, Mallinckrodt Medical).^32,144 Preparation of the ^186/188Re analogue of the precursor using the same method, however, resulted in very low yields.¹⁴⁵ Schibli et al. subsequently developed two kit procedures for the preparation of ¹⁸⁸Re tricarbonyl in yields greater than 85%, both of which require the reduction of perrhenate by the addition of the stronger reducting agent, H₃B·NH₃.¹⁴⁵ Park et al. eventually produced [¹⁸⁸Re][Re(OH₂)₃(CO)₃]⁺ in more than 97% yield in 2006 using Schibli's kit with the incorporation of borohydride exchange resin as an additional reducing agent and anion scavenger.¹⁴⁶ Although the Re(i) core is unstable to slow oxidation to perrhenate, it is made more stable by the coordination to a variety of organic ligands.^23,145 These ligands are able to be introduced via dissociative exchange with the three labile water molecules on the precursor, while the three facially-coordinated CO ligands are tightly held.^16,32,147 Monodentate, bidentate and tridentate ligands have been used to create stable complexes with the rhenium(i) tricarbonyl core,^{142,148–150} with complexation of the core to a biomolecule-linked chelator being a favourable reaction in the field of nuclear medicine.²³

The “2 + 1”chelator approach for the [^186/188Re][Re(CO)₃]⁺ core

While monodentate ligands are robust and coordinate efficiently, they are not versatile and may coordinate more than one biomolecule to the metal centre.¹⁵¹ It is for this reason that monodentate ligands are often paired with bidentate chelators to occupy the other two coordination sites. This “2 + 1” approach of a combination of monodentate and bidentate ligands has been shown to afford complexes with high in vitro stability with a decent level of flexibility.¹⁵² There is also a good opportunity to design very versatile complexes as the pharmacokinetics can be fine-tuned using one ligand while the targeting ligand may be incorporated with the other. While this approach has been used extensively in the preparation of ^99mTc-labelled biomolecules, it has less commonly been used to produce “2 + 1” complexes of radiorhenium. Since radiorhenium complexes require harsher radiosynthesis conditions, their synthesis would benefit from this “2 + 1” approach, as the biomolecule-bearing monodentate ligand could be incorporated in the final step.¹⁶ Completion of the coordination sphere of the [Re(CO)₃]⁺ core has been achieved in the past using monoanionic bidentate ligands such as 8-mercaptoquinoline (NS),¹⁵³ dithiocarbamate (SS),¹⁵⁴ 2-picolinic acid (NO),¹⁵⁵ hydroxyquinoline (NO)¹⁵⁰ and also 2-(hydroxyphenyl)diphenylphosphine (PO).¹⁵⁶ The monodentate ligand is usually neutral, with strong nucleophiles such as phosphines, imidazoles and isocyanides being used to ensure the ligand is inert to in vivo exchange.¹⁶ Although most of the reported chelator complexes bearing the [Re(CO)₃]⁺ core have been made with ^natRe as a means to characterise their ^99mTc counterparts, recent literature has shown success using ^186/188Re. Shegani et al. in 2021 synthesised a series of “2 + 1” tricarbonyl complexes with the general formula fac-[Re/^99mTc/¹⁸⁶Re(CO)₃(DDTC)(L)] which contained a bidentate diethyldithiocarbamate (DDTC) ligand and a variety of monodentate ligands (L).¹⁵² One of the [¹⁸⁶Re]Re-tricarbonyl complexes evaluated by this group is shown in Fig. 23a. Due to the slower kinetics of rhenium during radiolabelling, the [¹⁸⁶Re]Re-complexes were obtained in lower radiochemical yields than their ^99mTc equivalents.¹⁵² They did, however, remain stable at room temperature over 48 hours and showed high resistance to transchelation in vitro during histidine and cysteine challenge assays (>95% complex intact).¹⁵² These dithiocarbamate-containing [¹⁸⁶Re]Re-complexes with isocyanide, phosphine and arsine monodentate ligands also provide the opportunity for biomolecule conjugation, as the ligands are amenable to derivatisation. In particular, these highly lipophilic complexes could be conjugated with biomolecules which require the penetration of the cell membrane to reach their desired target.¹⁵²

Fig. 23. Structures of (a) [¹⁸⁶Re]Re(CO)₃(DDTC)(PPh₃)¹⁵² and (b) [¹⁸⁸Re]Re(CO)₃-tamoxifen.¹⁵⁷.

Another recent example in the search for radiorhenium-tricarbonyl complexes for use in targeted radionuclide therapy is the [¹⁸⁸Re]Re-tricarbonyl tamoxifen compound synthesised by Chhabra et al. in 2021 (Fig. 23b).¹⁵⁷ Tamoxifen is a selective estrogen receptor modulator that binds to the over-expressed estrogen receptors of breast cancer cells and inhibits their growth. Unfortunately, it is also a dose-dependent agonist of estrogen in the uterus, often leading to the development of uterine cancer.¹⁵⁷ This is why there is a need for a low-dose, targeted therapy using tamoxifen for the treatment of estrogen receptor-expressing breast cancer. Since Chhabra et al. found positive results with the non-invasive imaging of estrogen receptor-expressing breast cancer cells with [^99mTc]Tc-tricarbonyl, the same lab then explored [¹⁸⁸Re]Re-tricarbonyl tamoxifen for targeted therapy.^157,158 The prepared [¹⁸⁸Re]Re-tricarbonyl tamoxifen complex shown in Fig. 23b had a radiochemical yield of greater than 95% and only micro- to nanomolar amounts were required to target the breast cancer estrogen receptors while sparing healthy, nonestrogen receptor-expressing cells.¹⁵⁷

Tridentate bifunctional chelators for the [^186/188Re][Re(CO)₃]⁺ core

While the “2 + 1” combination of bidentate and monodentate ligands for the fac-[M(CO)₃]⁺ core does provide a unique versatility to the design of the resultant complex, this approach does not display the same degree of stability as the use of tridentate chelators. Tridentate chelators have received the greatest attention of all because they can be radiolabeled very efficiently and the chelate effect imparts a high thermodynamic stability on the overall complex in vitro and in vivo.^151,159 Since the use of tridentate ligands with the [M(CO)₃]⁺ core leaves no vacant positions available in the coordination sphere, these complexes do not interact with biomolecules in vivo and therefore exhibit fast tissue and blood clearance.¹⁵⁹ Of the tridentate chelators for the [M(CO)₃]⁺ core reported in literature, those containing N-, S- or O- donor atoms have been the most attractive, with ligands bearing at least one aromatic amine in particular being the most ideal for improved complex stability and reaction rates.^148,160 As is usually the case, most research done on bifunctional, tridentate ligands for the [Re(CO)₃]⁺ core is performed as a structural investigation for their ^99mTc analogues,^143,161,162 however, there are a number of examples making use of “hot” rhenium for targeted radiotherapy. These chelators include functionalised picolylamine derivatives,^163–166 histidine derivatives,^167,168 aminodiacetate¹⁶⁹ and triazoles via the “click-to-chelate” approach.^147,170 The “click-to-chelate” approach (Fig. 24) is an interesting and efficient strategy which allows the one-pot formation of the chelating system and biomolecule conjugation in a single step with high yields.^147,171 Wang et al. in 2012 was the first to use this method for developing triazole-based chelators for radiorhenium.¹⁷⁰ They radiolabelled the prepared triazole chelator with [¹⁸⁸Re]Re-tricarbonyl in 80% yield and showed that the triazole chelator made via click chemistry is an “extraordinarily ideal” chelator for radiorhenium.¹⁷⁰ Eychenne et al. developed this further in 2017 by radiolabelling two functionalised N₂O tridentate click ligands with [¹⁸⁸Re]Re-tricarbonyl.¹⁷² This showed that the “click-to-chelate” protocol is promising for developing BFCs for target-specific radiotherapy.

Fig. 24. The “click-to-chelate” approach for the one-pot synthesis of functionalised chelators for the [M(CO)₃]⁺ core.

While the majority of BFCs for radiorhenium make use of acyclic tridentate ligands, a number of studies have shown that macrocyclic triaza-containing (NNN) compounds are promising BFCs for radiolabeling biomolecules. These include NOTA and NODAGA – polycarboxylate derivatives of 1,4,7-triazacyclononane which can coordinate to the fac-[M(CO)₃]⁺ core, resulting in stable, hydrophilic complexes.^173,174 In 2018, Makris et al. first explored [^99mTc]Tc/[¹⁸⁶Re]Re-tricarbonyl core with NOTA and NODAGA BFCs conjugated to sst₂-ANT, the somatostatin receptor antagonist 4-NO₂-Phe-c(DCys-Tyr-DTrp-Lys-Thr-Cys)-DTyr-NH₂.¹⁷⁵ They were prepared in high radiochemical yields and biodistribution studies in healthy mice with the [^99mTc]Tc(CO)₃(NODAGA)-peptide bioconjugate revealed outstanding pharmacokinetics, such as negligible non-specific organ uptake and rapid renal clearance after 1 hour.¹⁷⁵ The next year, the same group further evaluated the radiochemistry and in vitro properties of the fac-[M(CO)₃]⁺ complexes, [¹⁸⁶Re]Re-NODAGA-sst₂-ANT and [^99mTc]Tc/[¹⁸⁶Re]Re-NOTA-sst₂-ANT, the synthesis of which is shown in Fig. 25.¹⁷⁶ Histidine and cysteine challenge experiments showed the radiolabelled complexes remained stable (≥93%) over 48 hours and rat serum stability experiments also resulted in low non-specific protein binding.¹⁷⁶ Although preclinical studies were only performed with the [^99mTc]Tc-NOTA/NODAGA-sst₂-ANT complexes in AR42J tumour-bearing mice, the results confirmed the potential of NOTA and NODAGA bioconjugates of both the [^99mTc]Tc- and [¹⁸⁶Re]Re-tricarbonyl cores for the development of theranostic radiopharmaceuticals.¹⁷⁶ In 2021, Makris et al. expanded on the applicability of their previous research by developing gastrin-releasing peptide receptor (GRPR)-targeting complexes of [^99mTc] and [^186/natRe]Re with NOTA and NODAGA BFCs.¹⁷⁷ The chelators were conjugated via a 6-aminohexanoic acid linker to the GRPR-targeting peptide antagonist, RM26 (_DPhe-Gln-Trp-Ala-Val-Gly-His-Sta-Leu-NH₂) (Fig. 25).¹⁷⁷ Again, the complexes displayed excellent in vitro and in vivo stability as well as good hydrophilicity and receptor-mediated tumour uptake.¹⁷⁷ While these results do show promise, they concluded that further modifications will be required to optimise the tumour-retention ability of the [¹⁸⁶Re]Re-bioconjugates in particular, thereby improving the theranostic potential of this approach.

Fig. 25. Proposed structures of the M(CO)₃-NOTA/NODAGA-peptide complexes where M = ^natRe, ^99mTc, ¹⁸⁶Re and BM = sst2-ANT¹⁷⁶ or 6-Ahx-RM26.¹⁷⁷.

Bifunctional chelators for ^186/188Re(vii)

The Re(vii) trioxo core

The [MO₃]⁺ core containing Re⁷⁺ or Tc⁷⁺ ions is not new, however, the stabilisation of this core with various chelating ligands has recently attracted interest for the development of pharmaceuticals. This is because the metal-trioxo core is small, highly hydrophilic and results in very stable complexes when coordinated to various chelators under mild reaction conditions.^178,179 Since the metal is already in its highest oxidation state, further oxidation in vivo is impossible. This is particularly important for the Re(vii) core, and therefore also radiorhenium(vii) complexes, as rhenium is more stable to reduction than technetium.²⁵ The [MO₃]⁺ core is best stabilised using tridentate ligands, with cyclic, N-donor ligands being superior to acyclic ligands in producing water and air-stable rhenium complexes.¹⁷⁸ Popular examples of these tridentate ligands include 1,4,7-triazacyclononane (tacn), 1,3,5-triazacyclohexane (tach) and their derivatives, as well as “scorpionate” ligands.^178,180,181 These ligands are aptly named by the similarity they bear to a scorpion, as two donor sites act like the “pincers” which coordinate to the metal and the third donor site acts as the “tail” that reaches over and stings the metal. Scorpionate ligands are particularly attractive as they can be easily fine-tuned to incorporate biomolecules, particularly on the central carbon or boron atom, as shown in Fig. 26.^{178,180,182–184} A number of scorpionate ligands have been reported to stabilise Re and ^99mTc cores, with complexation occurring in aqueous media and therefore showing compatibility with biomolecule labelling.^182,184 Although no radiorhenium(vii) complexes utilising the trioxo core have been reported to date, the success seen with cold rhenium and ^99mTc-trioxo-labelled biomolecules bears witness to the potential of radiorhenium for the development of targeted radiopharmaceuticals in the future.

Fig. 26. Example of a scorpionate ligand for rhenium with the incorporation of the biomolecule (BM) on the C or B central atom.

Finally, another new and interesting strategy for the radiolabelling of biomolecules using the M(vii) (M = Re, ^99mTc) core is the [3 + 2]-cycloaddition of trioxo complexes with olefins.^179,185,186 This results in the formation of a M(v)-glycolato complex. While this method of radiolabelling does make use of the BFC approach, the reaction in this case is ligand-centred (reaction takes place on the oxo ligands) as opposed to metal-centred which is conventional.¹⁸⁵ This method of complex-formation has been proven to work with ^99mTc, however, with Re, the diolate often cleaves upon heating, resulting in the release of the olefin. This occurrence was first encountered by Pearlstein and Davison in 1988 and continues to be an issue with [3 + 2]-cycloaddition-type reactions on a rhenium core today as the reverse reaction is favoured.¹⁸⁷ Encouragingly, Middleditch et al. in 2007 showed this not to be a problem when they reported the synthesis of a series of air-stable [3 + 2]-cycloaddition products from the reaction of rhenium oxo complexes with diphenyl ketene.¹⁸⁸ This indicates that there might be a possibility that, under certain conditions and with certain olefins, the [3 + 2]-cycloaddition reaction with rhenium may take place and present a promising way of incorporating a biomolecule into the final complex. The biomolecule can be conjugated to the alkene, with the choice of alkene determining the rates of labelling. Additionally, the backbone of the tridentate ligand may also be functionalised, allowing for the formation of a complex with dual targeting properties, as shown in Fig. 27.^179,185 Again, these reactions have only been reported for ^99mTc and “cold” Re reference complexes, however, their novel way of radiolabelling could provide an innovative and novel approach to designing BFCs for radiotherapy with radiorhenium.

Fig. 27. The [3 + 2]-cycloaddition approach to making Re-complexes with dual targeting properties.

Approaches to radiolabelling biomolecules with radiorhenium

A successful radiopharmaceutical for targeted radiotherapy comprises three parts: the radionuclide, the bifunctional chelator and the targeting vector. The first two components have been discussed in detail in the preceding sections. With focus on BFCs coordinated to radiorhenium, in this section, the various targeting vectors and bioconjugation strategies will be explored.

Types of labelling

Biomolecules may be bound to a radionuclide either directly or indirectly, depending on the application of the study. While direct labelling is efficient and has been used to radiolabel antibodies with ¹⁸⁸Re in the past, it is restricted to biomolecules with cleavable disulphide bonds which may serve as chelators for the radionuclide.^12,189,190 Although more complicated and time-consuming, the indirect labelling approach is often a better method of producing site-specific, radiolabelled biomolecules. This makes use of a BFC and has been shown to be a versatile way of radiolabelling biomolecules with radiorhenium. Indirect labelling can be further subdivided into pre-conjugation labelling (prelabelling) and post-conjugation labelling (postlabelling), each having their own advantages and disadvantages (Fig. 28).¹⁹¹

Fig. 28. The pre-conjugation and post-conjugation labelling approaches (image created with BioRender.com).

In the prelabelling method, also known as the preformed chelate approach, the BFC is complexed with the radiorhenium and is, thereafter, conjugated to a biomolecule.^12,189,190 This approach involves well-defined chemistry and has been used many times with various BFCs to radiolabel heat-sensitive biomolecules with radiorhenium, as highlighted in a review by Liu et al.²³ The shortfall in this method is the need for postlabelling purification – a time-consuming process of separating the radiolabelled chelator from any unbound chelator.^189–191

The post-conjugation labelling approach, where the BFC is first coordinated to a biomolecule and then radiolabelled, would ideally be a safer and quicker way of producing the desired radiolabelled conjugate.^12,189–191 No radioactive purification step is needed if the radiolabelling yield is >95%, leading to shorter radiation exposure times and no loss of the resulting radiopharmaceutical as a result of the purification process. This method is not practical for sensitive biomolecules, however, due to the harsh reaction conditions (e.g. high temperatures, non-physiological pH, etc.) required for radiolabelling. It is for this reason that the postlabelling approach has not been very popular, although several examples of radiorhenium-labelled biomolecule conjugates exist having been produced by this method.^59,105 Interestingly, Safavy et al. showed that trisuccin-antibody conjugates may be radiolabelled with ¹⁸⁸Re by incubation at 45 °C for 45 minutes, resulting in radiolabelled conjugates in high yield and with high in vivo stability.¹⁹² More recently in 2017, Kan et al. also showed that a MAG₃-RM26 peptide conjugate could be radiolabelled with ¹⁸⁸Re at room temperature within 1 hour.⁶⁸ These results show that the post-conjugation labelling method is promising, however, time will need to be taken to optimise the radiolabelling conditions in the presence of different biomolecules.

Bioconjugation strategies

An important factor to consider is the placement of the targeting vector in relation to the radionuclide so as to prevent the interference of one side with the functioning of the other.¹⁹³ To this end, the chemoselectivity and site-specificity of the modification reaction are essential in creating homogenous radiolabelled bioconjugates.^191,194 The chemoselectivity refers to the ability of the targeting vector to react preferentially at one reactive functional group over another while site-specificity denotes the ability of a biomolecule to be functionalised at a single position.^191,194 Amongst a sea of reactive amines, thiols, carboxylic acids and alcohols located at different positions on biomolecules, the latter often proves to be the most challenging task.¹⁹⁵ Random conjugation of the BFC to various sites on the biomolecule can affect the biological activity and pharmacokinetics of the resulting conjugate, therefore it is vital that precise control over bioconjugation may be attained for the effective clinical use of the radiopharmaceutical.^191,194

The majority of bioconjugation strategies make use of the side chains of lysine (–NH₂) and cysteine (–SH) residues on biomolecules, with the cysteine free thiols having a lower abundance in biomolecules and making their use more site-selective.¹⁹⁵ The popular reaction of the lysine primary amines with carboxylic acids or activated carboxyl groups or with an isothiocyanate on the BFC results in stable amide or thiourea bonds, as shown in Fig. 29A and B. A thioether bond is formed upon the Michael addition of a cysteine free thiol with a maleimide, or a derivative thereof, occurring under physiological conditions (Fig. 29C).^193,196 Unfortunately, it has been observed that the thioether bond formed is reversible in vivo, undergoing rapid retro-Michael addition reactions with thiol-containing biological molecules in plasma.^197,198 Bernardim et al. have therefore proposed the use of the carbonylacrylic functional group for chemoselective and irreversible cysteine bioconjugation and this seems to be a promising way of producing radiobioconjugates for in vivo applications in the future.^199,200

Fig. 29. Bioconjugation bond-formation strategies.

Finally, “click chemistry” also provides a way of conjugating biomolecules to chelators via highly selective and fast cycloaddition reactions.^191,201,202 The good site-specificity of this reaction stems from the use of bioorthogonal functional groups, which do not interact with biological moieties in vivo, nor with the final product.²⁰¹ Two excellent, up-to-date reviews by Rigolot et al. and Kaur et al. highlight some of the biorthogonal click chemistry methods used to target biomolecules.^203,204 The most well-known method is the copper-catalysed azide-alkyne cycloaddition (CuAAC) which forms a 1,4-disubstituted triazole, as shown in Fig. 29D. While this type of click reaction is high-yielding and results in a stable 1,2,3-triazole linker, it is often limited by the toxicity of the Cu(i) catalyst in vivo.^191,201 The presence of copper cations can also interfere with the functioning of a chelator, meaning that this bioconjugation method is unlikely to be employed for radiopharmaceutical applications.^191,201,205 To this end, strain-promoted azide–alkyne cycloaddition (SPAAC) reactions came about as a copper-free way of joining a targeting vector to a radionuclide (Fig. 29E).^203,206 By making use of an azide and a strained cyclooctyne, SPAAC reactions can occur in physiological conditions with relatively quick kinetics, making them interesting for use in pretargeting strategies in radioimaging and radiotherapy.^207–209 Finally, inverse electron-demand Diels–Alder (IEDDA) reactions are the latest copper-free addition to the biorthogonal click chemistry family, having the fastest in vivo reaction kinetics to date (Fig. 29F).^210,211 This reaction takes place between an electron-deficient diene, such as a tetrazine, and an electron-rich dienophile, such as a trans-cyclooctene and results in the formation of a stable pyridazine with the release of N₂.²⁰² The popularity and use of this highly-selective biocompatible reaction has only taken off over the past 5 years, however, it has already been used in a number of radiolabelling applications and pretargeting strategies with biomolecules.^212,213 With click chemistry showing significant advantages over other bioconjugation strategies, its application in radiolabelling biomolecules may revolutionise the development of radiopharmaceuticals for disease imaging and treatment.

Choice of targeting vector

Cancer cells are particularly well-suited to detection and treatment by targeted radiopharmaceuticals. This is because the cell surface-receptors that are usually exhibited in normal cells are overexpressed in cancer cells, allowing the targeting vector to have a much higher probability of binding to cancer cells than healthy tissue.¹⁹³ A number of biomolecules have been used to target these receptors, with the choice of biomolecule being based on properties such as its biological half-life and molecular size. It is imperative that the biological half-life of the biomolecule is compatible with the physical half-life of the radiometal to ensure the patient gets the optimum radiation-absorbed dose by the radionuclide before it is excreted from the body. The circulation time of the biomolecule in vivo is directly proportional to its molecular size.^193,214 Larger molecules with molecular weights above the glomerular filtration threshold (∼70 kDa) will remain in circulation in the body much longer than smaller molecules, which are eliminated rapidly via the kidneys.²¹⁴ Since ¹⁸⁶Re has a longer half-life (t_1/2 = 3.8 days), it would be more compatible with larger biomolecules such as antibodies, while ¹⁸⁸Re with the shorter half-life (t_1/2 = 17.0 hours) would be more well-suited to targeting vectors such as various antibody fragments and peptides.³⁴ A variety of targeting vectors have been used for targeted radiotherapy with radiorhenium, some of which have already been mentioned in previous sections of this review. These have been most commonly antibodies, antibody fragments, peptides, small molecules and even nanoparticles.^193,215

Antibodies are heavy, Y-shaped proteins with a high molecular weight of roughly 150 kDa.²¹⁶ Monoclonal antibodies (mAbs) are designed to bind to only one target, meaning that they can be widely utilised in the therapy of various cancers, although, since their discovery, their use has not been met with the expected popularity.¹⁹³ This is mostly due to their large size, often leading to reduced tumour accumulation and long circulation times of 3–4 weeks.^193,216 The ¹⁸⁶Re radioisotope has been used on a number of occasions to radiolabel mAbs using a BFC like N₃S MAG₃,^42,73,217 N₂S₂ MAMA²¹⁸ or N₂S₄,²¹⁹ with ¹⁸⁶Re-MAG₃-mAb in particular showing promising results in preclinical studies.²²⁰ In spite of it being less well-suited for radiotherapy with mAbs, several antibodies have been radiolabelled with ¹⁸⁸Re and tested preclinically in various tumours.^{34,89,189,221–223} Antibody fragments have been explored as an alternative targeting vector to full monoclonal antibodies and are deemed to have more favourable pharmacokinetics.^215,224 They are theoretically able to retain the strong target-binding affinity of full mAbs, but their smaller size means they have faster blood-clearance via the kidneys with low liver accumulation.²¹⁵ Radiolabelling of antibody fragments has been explored with radiorhenium,²²⁵ with the scope for this method of tumour targeting set to expand to in the same way as seen with ^99mTc.^226–228 Nanobodies, or single domain antibodies, are particularly interesting as emerging vectors for cancer applications.^229,230

Similar to antibody fragments, peptides have favourable properties that make them desirable as targeting agents for radiotherapy. In addition to rapid target tissue penetration via receptor-mediated internalisation, they exhibit fast blood and non-target tissue clearance and elicit a low immune response.^190,231 Chemically speaking, they are also easy to synthesise and are more robust against the harsh conditions required for radiolabelling.¹⁹⁰ The combination of their advantageous biological and chemical properties makes peptides ideal targeting vectors for radiorhenium labelling, particularly with the shorter-lived ¹⁸⁸Re.^22,34,231 Peptides that have been radiolabelled with radiorhenium include somatostatin analogues,^{176,232–236} bombesin analogues^38,237,238 and others such as those targeting melanoma.^239–241 Currently, ¹⁸⁸Re-P2045 targeting SSTR-positive lung cancer is the only ¹⁸⁸Re-labelled peptide to have reached clinical trials.²⁴² While the initial results of phase I clinical trials were promising, they found that the renal accumulation of the radiolabelled peptide was too high and could lead to renal toxicity, causing the trial to be halted.²⁴² Nonetheless, new strategies continue to be explored for designing and improving radiorhenium-labelled peptides for targeted radiotherapy. Small molecules and even nanoparticles also offer an interesting way of targeting cancer cells and have the advantage of being more robust than the heat- and pH-sensitive biomolecules. Several examples of radiorhenium-labelled small molecules and nanoparticles for radionuclide therapy exist and appear to be a promising method of improving the diagnosis and treatment of tumours.^{44,163,243–246}

Future outlook – where to from here?

As our understanding of the complexities of cancer biology continues to expand, so too does our interest in targeted radionuclide therapy. A major contributor to this has been progress in the “see and treat” approach of theranostics, which is viewed as the “marriage” of diagnostic imaging and targeted therapy.²⁴⁷ By this definition, one of the most studied and reviewed “couples” in this marriage is that of ^99mTc and ^186/188Re. With ^99mTc being used for roughly 90% of all SPECT imaging applications, it is imperative that we continue to develop our understanding on the therapeutic potential of radiorhenium. Although both isotopes of rhenium have attractive physical and chemical properties, the use of ¹⁸⁸Re is of particular interest due to its ease-of-production from a ¹⁸⁸W/¹⁸⁸Re generator.^47,248 An obstacle to achieving this, however, has often been the availability of the adequate quantity and specific activity of the parent radionuclide, ¹⁸⁸W. This can only be achieved in very high flux reactors, only three of which exist in the world (HFIR at Oak Ridge in the USA, the SM3 reactor in Dimitrovgrad, Russia and BR2 in Mol, Belgium).²⁴⁸ Additionally, even if sufficient ¹⁸⁸W to meet clinical needs could be produced by these reactors, there is a lack of sites available to process the irradiated targets for use in the ¹⁸⁸W/¹⁸⁸Re generator. This has placed a long-time handicap on the development of potential ¹⁸⁸Re radiopharmaceuticals, however, thanks to the efforts of Dr Ajit Padhy and the International Atomic Energy Agency (IAEA), the widespread use of the ¹⁸⁸W/¹⁸⁸Re generator became a possibility in the early 2000s.^249,250188Re for radiopharmaceutical applications could now conveniently be produced on-site in hospitals, although the cost-effectiveness of the generator use in clinical settings still remains a problem.⁴⁶ Often, lone institutions are not pursuing enough nuclear medical applications to optimally use the ¹⁸⁸Re eluted from the generator, making the cost per patient extremely high. A way to improve the cost-effectiveness of the ¹⁸⁸W/¹⁸⁸Re generator in the future could be to establish centralised radiopharmacy units specifically for the administration of ¹⁸⁸Re radiopharmaceuticals, where the generator may be used for a variety of applications and not just single-use studies.⁴⁷ These centres could be set up within a close-enough proximity to both hospitals and research institutions, thereby helping with both patient treatment and facilitating novel ¹⁸⁸Re radiopharmaceutical research.

With regards to ^186/188Re radiopharmaceuticals, BFCs that can be used to radiolabel biomolecules with radiorhenium in a quick and stable manner are still urgently needed. As highlighted in this review, a lot of progress has already been made towards developing novel metal cores of ^186/188Re in various oxidation states, each of which has its own pros and cons. The ^186/188Re(v) core has been the most extensively studied, with the well-known MAG₃ chelator forming stable rhenium complexes with limited reoxidation to perrhenate after chelation. Its downfall is that high temperatures are required in the radiolabelling process, making it unsuitable for radiolabelling heat-sensitive biomolecules. Future work in this regard needs to focus on finding a way to get the “best of both” when it comes to radiolabelling a biomolecule with radiorhenium, by incorporating the strong coordination capabilities of N₃S and N₂S₂ chelators with heat-sensitive targeting biomolecules. To this end, the pretargeting strategy of radionuclide imaging and therapy could be an ideal solution as it, in theory, offers better therapeutic efficiency and lowers the received dose to healthy tissues, compared to conventional methods (Fig. 30).²⁵¹ Since the chelator part and the targeting part are produced and administered separately, consideration does not need to be given to finding biocompatible radiolabelling conditions. This would be perfect for treatment with radiorhenium, particularly the shorter-lived ¹⁸⁸Re isotope, as it involves allowing the targeting vector to accumulate in the tumour for a number of days before administering the fast-clearing radionuclide. Research into bioorthogonal bioconjugation reactions such as the copper-free click reactions, SPAAC and IEDDA, is expanding rapidly and these reactions have proven to have the ideal in vivo reaction kinetics for use in the pretargeted strategy.

Fig. 30. The pretargeting strategy in nuclear imaging and radiotherapy (image created with BioRender.com).

Over the past few years, more and more emphasis has been placed on biochemistry as opposed to inorganic chemistry when it comes to designing chelators, with a lot of focus being placed on the biomolecule conjugated to the metal chelate. The need for BFCs and metal cores that are water-soluble and stable in vivo has become more important, with the hypothesis that the smaller the complex, the lower the chance of altering the biological activity of the conjugate in vivo.²⁵² In this regard, the Re(i)-tricarbonyl core is becoming increasingly popular as its low oxidation state makes it chemically very inert and therefore quite stable for biomedical applications. Its “closed” octahedral coordination sphere also offers protection to reoxidation – a consequence often observed with the more “open” Re(v)-oxo species.²⁵² Since reoxidation to perrhenate has been the biggest obstacle in the design of effective rhenium radiopharmaceuticals, it would be no surprise to see more of the Re(i)-tricarbonyl type complexes dominating the rhenium radiopharmaceutical research field in the future. The one-pot “click-to-chelate” method of producing functionalised chelators for the [M(CO)₃]⁺ core may also be an interesting approach to use going forwards. This allows the bioconjugation and radiolabelling to happen in one step and could be a practical and efficient way to radiolabel heat-sensitive biomolecules, with the triazole formed being extremely stable for radiorhenium coordination.

Further than the design of suitable bifunctional chelators for radiorhenium, other groups are working on improving the overall tumouricidal effect of the ionising radiation produced by radiorhenium. Pourhabib et al. has shown that using a combination of ¹⁸⁶Re and ¹⁸⁸Re could produce complementary radiopharmaceuticals that may be administered together as a mixture.²⁵³ They determined that using the blend of the radiorhenium isotopes could be more effective at targeting tumours of various sizes rather than utilising each isotope independently.²⁵³ This could be a very interesting way of ensuring the best outcome to treat tumours, especially since it was determined that the amounts of impurities in the simultaneous production of ¹⁸⁶Re and ¹⁸⁸Re via irradiation of ^natRe were negligible.

In spite of the past challenges faced in the design of novel rhenium radiopharmaceuticals, rhenium still continues to be an attractive radionuclide for use in nuclear medicine. In addition to making the ¹⁸⁸W/¹⁸⁸Re generator more cost-effective, there is also a great need for interdisciplinary thinking going forward in order to restore the popularity of radiorhenium to what it once was. No longer can we work on developing rhenium radiopharmaceuticals as isolated physicists, radiochemists, inorganic synthetic chemists or biologists. It will take a merging of these fields to fully grasp the complexities of radiopharmaceutical design, for it is in unity that ultimate success may be achieved.

Conflicts of interest

The authors declare no competing financial interest.

Biographies

Biography

Diana R. Melis.

Diana Melis obtained her B.Sc. in 2016 and B.Sc. Hons. in 2017 from the University of Cape Town (South Africa). Following on from that, she obtained her M.Sc. in bioorganometallic chemistry in 2019 at the same university under the supervision of Assoc. Prof. Gregory Smith. Diana began her PhD in 2021 with SCK CEN (Mol, Belgium) and Chimie ParisTech, PSL University (Paris, France) under the supervision of Prof. Gilles Gasser and Dr Maarten Ooms. Her research focuses on developing novel, water-soluble chelators for ¹⁸⁸Re-radiopharmaceuticals.

Biography

Andrew R. Burgoyne.

Andrew R. Burgoyne received his Ph.D. from the University of Cape Town (South Africa). After a postdoctoral stint at the College of Pharmacy of the University of Hawaii in Hilo (HI, USA), he joined the Belgian Nuclear Research Centre SCK CEN (Mol, Belgium) to develop new chelators for radiopharmaceuticals and became heavily involved in radioisotope purification. Recently, Andrew moved to Oak Ridge National Laboratory (TN, USA) to continue focusing on radioisotope production.

Biography

Maarten Ooms.

Maarten Ooms obtained his PhD in Pharmaceutical Sciences from the KU Leuven, Belgium followed by a postdoctoral appointment at the Molecular Imaging Branch of the National Institutes of Mental Health (NIMH, Bethesda, MD). Currently, he is working at the Belgian Nuclear Research Center (SCK CEN) as a project leader responsible for the radiopharmaceutical research within the institute.

Biography

Gilles Gasser.

Gilles Gasser started his independent scientific career at the University of Zurich (Switzerland) in 2010 before moving to Chimie ParisTech, PSL University (Paris, France) in 2016 to take a PSL Chair of Excellence. Gilles was the recipient of several fellowships and awards including the Alfred Werner Award from the Swiss Chemical Society, an ERC Consolidator Grant, the European BioInorganic Chemistry (EuroBIC) medal and the Pierre Fabre Award for therapeutic innovation from the French Société de Chimie Thérapeutique. Gilles' research interests lay in the use of metal complexes in different areas of medicinal and biological chemistry.