An overview of the role of radionuclides in targeted cancer treatment: application of biomarkers for patient selection and developments to improve treatment efficacy

Abstract

Radiopharmaceuticals for targeted radionuclide therapy (TRT) of tumours consist of a radionuclide conjugated to a component that can target the cancer. Several TRT radiopharmaceuticals have been licensed for the treatment of lymphoma, neuroendocrine and prostate cancers. The outcomes from two TRT trials, NETTER for neuroendocrine and VISION for prostate cancer, demonstrated beneficial outcomes. These findings have increased interest in the application of TRT in the treatment of prostate cancer and expansion to other cancer types. Patient selection for TRT is based on a measure of the overexpression of a target receptor on the cancer. To facilitate this, imaging is carried out using a similar targeting moiety to that used for treatment but labelled with an imaging radionuclide. Theragnostic pairs are selected to enable imaging and treatment with the same construct providing accurate predictions of the pharmacokinetics of the therapeutic in patients. This review covers the imaging biomarkers that act as companion diagnostics for TRT pharmaceuticals and the development of radiopharmaceuticals targeting other cancer types enabling expansion of TRT to these cancers. These include strategies to target cancer cells specifically and a pan-cancer approach by targeting fibroblast-activated protein (FAP) upregulated on cancer-associated fibroblasts (CAF). FAP-targeted radiopharmaceuticals are useful for diagnosis and staging but have drawbacks for TRT. Approaches to improve the efficacy of TRT including the use of high linear energy transfer (LET) alpha-emitters and pre-targeting and combination treatments are also covered. As described in this review, not all patients benefit from TRT making the case for predictive biomarkers. This is particularly important for the more damaging alpha emitters.

Introduction

Treating cancers using radionuclides began in the 1940s with the use of radioiodine to treat hyperthyroidism and thyroid cancer, exploiting the affinity of thyroid tissue for iodine, a property maintained by some thyroid cancers (1). The β-emitter phosphorus-32 (³²P) has been used in the form of [³²P]H₃PO₄ for several decades to treat some blood disorders, including polycythaemia vera and bone pain from metastatic cancer (2). More recently, in 2013, [²²³Ra]RaCl₂, commercially known as Xofigo^®, was approved by the Food and Drug Administration (FDA) to treat bony metastasis in patients with advanced prostate cancer (3). Each of these applications uses the diseased tissue’s elemental affinity to concentrate radioactive versions of the element or their mimetics.

Many radiopharmaceuticals consist of complex constructs composed of a targeting moiety attached to a radionuclide via a carrier (see Figure 1 ). The structure of the carrier depends on the radionuclide. Medical radionuclides are often a metal and require chelation. Commonly used chelators include diethylenetriamine pentaacetic acid (DTPA) for technetium-99m (^99mTc) and bismuth-213 (²¹³Bi) (4); deferoxamine (DFO) for zirconium-89 (⁸⁹Zr); macrocyclic chelators including DOTA for the trivalent radiometals such as gallium-68 (⁶⁸Ga), scandium-44 (⁴⁴Sc), yttrium-90 (⁹⁰Y), lutetium-177 (¹⁷⁷Lu), and other radio-lanthanides including actinium-225 (²²⁵Ac); and 1,4,7-triazacyclononane-1,4,7-triacetic acid (NOTA) for Ga and copper (Cu) radionuclides (5).

Figure 1.

Diagrammatic representation of a targeted radionuclide therapy (TRT) radiopharmaceutical radiolabelled with metal radionuclides. Targeting groups include antibodies, truncated antibodies, and peptides. Examples of linkers include short linear molecules, polyethylene glycol (as monomer or polymer), and a nanoparticle. Metal radionuclides (as in the example in the diagram) are captured in a chelator conjugated to the linker. *Arginylglycylaspartic acid.

The targeting moiety can be a peptide, an antibody (Ab), or smaller-sized engineered antibody mimetics. These small constructs include fragment antibody (scFv) and minibodies (Mb) (two scFv to increase avidity), and affibodies—small robust protein binders and nucleic acid-based structures called aptamers (6–8). Affibodies are approximately 7 kDa in size and have picomole (pM) target affinity (9). However, they can demonstrate high kidney uptake due to reabsorption by the renal tubules (9). Modifications to the amino acid sequence, hydrophilicity, surface charge, chelate, and radionuclide can improve tumour/non-tumour uptake ratios (9).

For this review, papers from the past 5 years (mostly) were selected from a “Web-of-Science” search using the following keywords: targeted radiotherapy, molecular radiotherapy, radioimmunotherapy, and cancer. The purpose of this review is to provide a background for non-specialists in these areas and a comprehensive survey of the current research and clinical state of targeted radionuclide therapy (TRT).

Radionuclides commonly used in medicine

Table 1 shows examples of radionuclides commonly used in nuclear medicine. Imaging radionuclides are γ-emitters for single-photon emission computer tomography (SPECT) or positron (β⁺) emitters for positron emission tomography (PET). Technetium-99m (^99mTc) is used in the majority (>80%) of nuclear medicine scans due to its practical half-life (t_1/2 = 6 h) and ideal imaging γ-emission energy of 140 keV, with high detection efficiency (10). Fluorine-18 (¹⁸F) is the most commonly used PET isotope which has a low positron energy (0.6 MeV) and produces high-resolution images (11). The positron-emission tomography (PET) radionuclides with short t_1/2s, fluorine-18 (¹⁸F) and gallium-68 (⁶⁸Ga) (t_1/2 of 110 min and 68 min, respectively), are suitable for labelling small ligands such as peptides, which demonstrate rapid (hours) blood clearance. Antibodies with longer blood residence times (days) are labelled with longer-lived radionuclides, particularly Zirconium-89 (⁸⁹Zr) with a t_1/2 of 78 h (12).

Table 1.

Examples of radionuclides used clinically with characteristics.

Emission	Range	LET	Examples (range in tissue - for β-emitters)	Use
γ			111In, 99mTc	Imaging
Positron (β+)			89Zr, 18F	Imaging
β	Up to 10 mm	low	177Lu (1mm), 90Y (10 mm)	TRT
α	Up to 100 μm	high	223Ra, 225Ac	TRT
Auger electrons	<0.5 µm	high	125I, 89Zr, 111In, 123I	TRT

Zr, zirconium; In, indium; Tc, technetium; Lu, lutetium; Y, yttrium; Ra, radium; Ac, actinium; I, iodine.

Patients selected for TRT have pretreatment SPECT or PET scans to ensure tumour expression of the target receptor. Therapy with radionuclides is delivered using β- or α- and less commonly Auger emitters (13). The deposition of energy within the tumour from radioactive decay causes DNA damage. When unrepaired or incorrectly repaired, DNA damage can result in cell death. High linear energy transfer (LET), characteristic of α- and Auger emissions, results in high-density localised DNA damage (13), whereas sparsely ionising but longer range β-emitters improve radiation dose heterogeneity across tumours (14).

Production of radionuclides

Most medical radioisotopes are produced in a nuclear reactor (e.g., ¹⁷⁷Lu), or through proton or ion particle bombardment in a cyclotron (15, 16). The imaging radionuclide ^99mTc is produced in generators from molybdenum-99 [⁹⁹Mo], delivered weekly to nuclear medicine departments (17). There is currently great interest in treating patients with neuroendocrine tumours (NENs) and other cancers using Actinium-225 (²²⁵Ac)-labelled radiopharmaceuticals (18). Actinium-225 is obtained as a decay product of ²²⁹Th of which there is a limited supply (enough to produce 68 GBq/year of ²²⁵Ac—just a few hundred patient doses) (19). Several production pathways attempting to overcome supply limitations of ²²⁹Th include neutron irradiation of ²²⁶Ra and accelerator routes (19). Table 2 shows fabrication routes for other clinically used radionuclides and ones with potential clinical application. Isotopes of terbium have great potential for imaging and therapy, but their production is problematic. The reader is referred to a comprehensive review on production routes for Tb radioisotopes (47).

Table 2.

Established and novel methods for production of radionuclides.

Radionuclide (emission)	Reaction	Cyclotron/neutrons/ Generator	Reference
Diagnostic radionuclides
Positron emitters
64Cu (β+, β− t1/2 = 12.7 h)	63Ni (p, γ) 64Cu 64Ni (p, n) 64Cu* 65Cu (p, pn) 64Cu 67Zn (p, α) 64Cu	Medium flux nuclear reactor Cyclotron Cyclotron 18 MeV Cyclotron 18 MeV	(20) (21) (21) (21)
48V (β+ t1/2 = 16 d)	NatTi (p, n) 48V	Cyclotron 18 MeV	(22)
43Sc (β+, EC t1/2 = 3.9 h)	42Ca (d, n) 43Sc NatCa (α, p) 43Sc	Cyclotron 18 MeV	(23) (24)
44Sc (β+ Eave 0.63 MeV t1/2 = 3.93 h)	44Ca (p, n) 44Sc	Cyclotron (15–25 MeV)	(25)
72As (β+ 2.47 MeV t1/2 = 26 h)	72Ge (p, n) 72As	Cyclotron	(26)
66Ga (β+ t1/2 = 9.49 h)	66Zn (p, n) 66Ga	Cyclotron	(27)
69Ge (β+, β− Emax = 1.2 MeV t1/2 = 39 h)	69Ga (p, n) 69Ge	Cyclotron	(28)
124I (γ, β+, t1/2 = 4.2 h)	124Te (p, n) 124I	Cyclotron	(29)
89Zr (β+, t1/2 = 3.3 d)	89Y (p, n) 89Zr	Cyclotron	(30)
SPECT radionuclides
123I (EC γ, t1/2 = 13.3 h)	127I (p, 5n) 123Xe, 123Xe decays to 123I	Cyclotron	(31)
111In (EC γ, t1/2 = 2.8 d)	112Cd (p, 2n) 111In	Cyclotron	(32)
99mTc (EC γ, t1/2 = 6 h)	Decay of 99Mo	Generator	(17)
Therapeutic radionuclides
90Y (β−, t1/2 = 64 h)	90Sr (β−) 90Y	Decay of 90Sr	(33)
177Lu (β−, γ, t1/2 = 6.65 d)	176Lu (n, γ) 177Lu	Nuclear reactor	(34)
223Ra (α, t1/2 = 11.4 d)	227Ac (α, β) 223Ra	Generator	(35)
213Bi (α, t1/2 = 46 min)		Generator; parent 225Ac	(36)
225Ac (α, t1/2 = 9.92 d)	Purified from 229Th 226Ra (p, 2n) 225Ac	Cyclotron 16 MeV	(19)
211At (α, EC t1/2 = 7.2 h)	209Bi (α, 2n) 211At	Cyclotron K150	(37)
32P (β−, t1/2 = 14.3 d) Emax1.7 MeV	32S (n, p) 32P	Neutron irradiation	(38)
153Sm (β− Emax=0.81MeV γ 103 keV t1/2 = 1.93 d)	152Sm (n, γ) 153Sm	Nuclear reactor	(39)
67Cu (β−, t1/2 = 61.8 h)	70Zn (p, α) 67Cu	Cyclotron 30 MeV	(40)
47Sc (β− E 0.6 MeV t1/2 = 3.35 d)	46Ca (n, γ)47Ca, 47Sc 47Ti (n, p)47Sc NatV (n, p)47Sc	Thermal neutrons High energy neutrons Cyclotron 20 MeV	(41) (41) (42)
170Tm (β− Emax= 0.97 MeV t1/2 = 128 d)	169Tm (n, γ) 170Tm	Thermal neutron flux	(43)
165Er (Auger t1/2 = 10.4 h)	165Ho (p, n) 165Er	Cyclotron 16 MeV	(44)
167Tm (Auger, γ208 KeV t1/2 = 9.25 d)	168Er (p, 2n) 167Tm	Cyclotron 23 MeV	(45)
195mPt (Auger t1/2 = 4 d)	195Pt	Neutron activation	(46)

¹Enriched ⁶⁴Ni is expensive.

Bi, bismuth; S, sulphur; P, phosphorus; Er, erbium; Ho, holmium; Ni, nickel; Cu, copper; Zn, zinc; Ga, gallium; Ge, germanium; At, astatine; V, vanadium; Ti, titanium; Sc, scandium; Ca, calcium; Ac, actinium; Sm, samarium; Tm, thulium; Pt, platinum; At, astatine; Te, tellurium.

Theragnostic systems: imaging and therapy with the same construct

Imaging biomarkers include measures of the uptake of single-photon emission computer tomography (SPECT) and positron emission tomography (PET) tracers based on diagnostic analogues of TRT therapeutics (48). The tissue/tumour uptake is usually defined as a standardised uptake value which is uptake per gram tissue per injected dose of the tracer. Tracers based on TRT molecules are used as part of a TRT inclusion criteria and for dosimetry prediction and dose tailoring of the TRT (48).

Theragnostic strategies involve imaging and treatment of the tumour using the same radiopharmaceutical (if the radionuclide has an imaging emission) or imaging with an otherwise identical radiopharmaceutical as for TRT but labelled with an imaging radioisotope (theragnostic pair) (49). Tumour uptake and biodistribution of the imaging radiopharmaceutical informs on the potential benefit of TRT for that patient (see Figure 2 ).

Figure 2.

(A) Scheme showing decision to treat with TRT (targeted radionuclide therapy) based on imaging. (B) Radiopharmaceutical targets cancer cells with radionuclide for imaging and therapy. The molecule for imaging and therapy is identical for a theragnostic except for the radionuclide.

Theragnostic systems are a rapidly growing component of nuclear medicine combining radionuclide interchangeability to facilitate imaging and therapeutic capability (49). In addition to confirming patient suitability for TRT, pretreatment imaging allows predictive dosimetry to tailor administered therapeutic dose (50). Radioiodine, which has a long history in nuclear medicine has a multitude of isotopes. The commonly used radioiodine nuclides for medical applications, ¹²³I, ¹²⁴I, ¹²⁵I, and ¹³¹I, between them achieve single-photon emission computed tomography (SPECT) imaging, positron emission tomography (PET) imaging from gamma, positron (β⁺) decay properties, Auger, and β-emission for targeted radiotherapy (51).

Many radionuclides are metals and require chelation to enable stable integration into a radiopharmaceutical. Some metals, for example, scandium (Sc), have multiple isotopes suitable for imaging and treatment (52). Thus, ⁴³Sc (t_1/2 = 3.9 h) and ⁴⁴Sc (t_1/2 = 4 h) are positron emitters whereas ⁴⁷Sc (t_1/2 = 3.35 d) is a β-emitter, suitable for therapy and a primary gamma-emission suitable for SPECT. Scandium forms stable complexes with the chelator, 1,4,7,10-tetraazacyclododecane-1,4,7,10-tetraacetic acid (DOTA), but complexation requires heating to 80°C (52). Terbium has four particularly valuable radionuclides for nuclear medicine ¹⁴⁹Tb (t_1/2 = 4.2 h, α, β⁺-emitter), ¹⁵²Tb (t_1/2 = 17.5 h, EC and β⁺ emitter), ¹⁵⁵Tb (t_1/2 = 5.3 d, EC SPECT), and ¹⁶¹Tb (t_1/2 = 6.9 d, β⁻, Auger) (47).

Chelators which allow coordination of chemically diverse metals are particularly useful, allowing complexation of radiometals with different emission characteristics for broad-range theragnostic systems. Simms et al. (53) have recently developed a chelator that can bind to [²²⁵Ac]Ac³⁺, [¹⁷⁷Lu]Lu³⁺, [¹¹¹In]In³⁺, or [⁴⁴Sc]Sc³⁺, enabling delivery of treatment with α- and β-emitters and imaging with SPECT and PET, respectively, on the same molecule.

Imaging ²²³Ra-SPECT generates poor images due to the small photon abundance and scarcity of the injected activity (54). Due to the similar co-ordination chemistry of Ra and Ba, it has been suggested that ²²³Ra could form a theragnostic pair with ¹³¹Ba (t_1/2 = 11.5 d) and ^135mBa (t_1/2 = 28.7 d). ¹³¹Ba and ^135mBa decay by electron capture with γ-emissions of 124 keV (30%) and 268 keV (16%), respectively. Suitable multidendate chelators are being produced (54). An alternative approach to chelation includes encapsulation of [²²³Ra] by hydroxyapatite nanoparticles (55) and/or linkage to nanoparticles (56).

The chelator DOTA can form stable chelates with cerium (Ce), Th, and Ac (57). Cerium-134 decays (t_1/2 = 3.2 days) to the positron emitter ¹³⁴La (t_1/2 = 6 min). The in vivo generator ¹³⁴Ce/¹³⁴La enables application of the short-lived ¹³⁴La as a PET imaging nuclide for both alpha-emitting ²²⁵Ac and ²²⁷Th radionuclides. MicroPET images of [¹³⁴Ce]Ce-DOTA-Trastuzumab demonstrated high tumour uptake and low bone and liver uptake analogous to previously reported [²²⁵Ac]Ac-DOTA-Trastuzumab biodistribution results (57).

Preclinical/early clinical TRT radiopharmaceuticals

Experimental approaches to treating prostate cancer with TRT, explored at the preclinical stage, include using different prostate-specific membrane antigen (PSMA) ligands, radionuclides, or targeting other receptors upregulated on prostate cancer cells (23, 58–60). Garnuszek et al. (58) radiolabelled PSMA-D4 with the three β-emitters ¹⁷⁷Lu, ⁹⁰Y, ⁴⁷Sc, and the α-emitter ²²⁵Ac. All radiocomplexes demonstrated high accumulation in LNCaP tumour xenografts and rapid clearance from blood and non-target tissues. Scandium-47 forms a theragnostic pair with the β⁺ -emitting ⁴³Sc. ⁴³Sc- and ⁴⁷Sc-PSMA-617 demonstrate congruous uptake by LNCaP-ENZaR xenografts (23). Other α-emitting radionuclides including, ²¹²Pb and thorium-227 (²²⁷Th) (t_1/2 = 18.7 days), and the radiohalogen, astatine-211 (²¹¹At), have been investigated in preclinical research targeting PSMA (59). Böhnke et al. (60) fabricated a PSMA ligand from PSMA-617 with a modified linker and chelator (carboxy-HOPO), which demonstrated stable chelation of ²²⁷Th, high uptake by PSMA-expressing tumours in mice, and fast renal clearance.

Prostate cancer overexpresses other cell surface proteins besides PSMA. These include the proteases human kallikrein peptidases (HK2 and HK3) and receptors such as delta-like ligand 3 (DLL-3), CD46, and CUB domain-containing protein 1 (CDCP1) (59). Prostate stem cell antigen (PSCA) which is overexpressed on metastatic prostate, pancreatic, and bladder cancer cells is an alternative target to PSMA (61). Antibodies conjugated to chelated metals are internalised on binding to surface receptors on both tumour and non-tumour cells and the metal retained in the lysosomal compartment. Tsai et al. (7) compared the anticancer efficacy and biodistribution of the anti-PSCA minibody, A11 Mb labelled by iodination with ¹³¹I, which is not retained, with a [¹⁷⁷Lu]-DTPA. Both exhibited similar cancer cell killing in vivo. However, dosimetry, determined using immunoPET studies with ¹²⁴I and ⁸⁹Zr as surrogates for ¹³¹I and ¹⁷⁷Lu, demonstrated clearance of ¹²⁴I from liver and kidneys but retention of ⁸⁹Zr by kidneys. This suggests that ¹³¹I is a better choice for delivering tumour-inhibitory radiation dose but minimising non-tumour retention (7). However, trans-iodination associated with halogen-labelled radiopharmaceuticals can result in non-tumour exposure.

Studies of peptides targeting the gastrin-releasing peptide receptor (GRPR), on the surface of localised and metastatic prostate cancers, are also underway (62). [⁶⁴Cu]Cu-SAR-BBN is in clinical development for PET imaging of GRPR-expressing prostate cancer using bombesin peptides. A preclinical study has shown that Cu-SAR-BBN labelled with the therapeutic radionuclide [⁶⁷Cu] is an effective treatment for PC-3 tumours (62). Radiolabeled GRPR antagonists are considered safer than agonists and have shown higher tumour uptake and clearance than agonists. Kanellopoulos et al. (63) labelled GRPR antagonists with ^99mTc radiotracers demonstrating high uptake by GRPR-expressing tumours.

Rapid clearance of small-molecule ligands such as [²²⁵Ac]Ac-PSMA-617 can limit tumour delivery (64). To increase circulatory residence time, albumin-binding entities can be incorporated into radiopharmaceuticals enabling binding to circulatory albumin. The benefit of this approach was examined in [²²⁵Ac]Ac-PSMA-ligands ([²²⁵Ac]Ac-SibuDAB) (64) and [⁶⁴Cu]Cu-PSMA-BCH (65). Both agents demonstrated increased circulation time and tumour uptake (64, 65). Response of xenografts bearing PSMA-expressing cancer cells was greater in mice injected with [²²⁵Ac]Ac-SibuDAB compared with [²²⁵Ac]Ac-PSMA-617 (64). Uptake by normal tissue was also greater but did not result in greater toxicity (64).

Dosimetry modelling can be highly informative and contribute to decisions regarding radionuclide of choice. A modelling study of the α-emitting ²²⁵Ac, ²¹¹At, ²¹²Pb, ²²³Ra, and ²²⁷Th and the β-emitting therapeutic radionuclides ⁶⁷Cu, ¹³¹I, ¹⁷⁷Lu, and ⁹⁰Y demonstrated that the β- and α-emitters ¹⁷⁷Lu and ²¹¹At respectively are most suited for prostate radionuclide therapy because they can reduce toxicity exposure to surrounding organs but provide sufficient dose to treat the prostate tumour (66).

Table 3 summarises many of the different proteins upregulated on other cancer types and the imaging and TRT radiopharmaceuticals for their targeting that are in preclinical and in some cases clinical trials.

Table 3.

Targets preclinically evaluated as TRT for different cancers.

Target	Cancer type	Ligand	Radionuclide	Reference
PD-1/PD-L1 (programmed death receptors)	Non-small cell lung carcinoma; melanoma; colorectal cancer (CRC)	PD-1	[68Ga]	(67)
Neurotensin receptor	CRC	Neurotensin	[68Ga], [177Lu]	(68)
Gonadotrophin-releasing factor receptor (GRFR)	GRFR-expressing cells, e.g., prostate, breast, ovarian	Leuprolide	[68Ga], [177Lu]	(69)
Insulin-like growth factor-1 (IGF1) receptor	Several cancer types	IGF1 receptor affibodies	[68Ga], [111In]	(70)
IGF2R receptor	Osteosarcoma	IGFR2-Fab1	[225Ac], [111In]	(71)
Glypican-3 (GPC3)	Hepatocellular carcinoma (HCC)	RAYZ8009 (macrocyclic peptide)	[177Lu], [225Ac]	(72)
	Hepatocellular carcinoma (HCC)	Anti-GPC3-Ab (αGPC3)	[227Th]	(73)
	Glioblastoma	GC332 Mab vs. scvf3	[225Ac] [89Zr]	(74)
Nectrin-1	Breast cancer	NP137 Mab	[111In], [177Lu]	(75)
CXCR4	Non-small cell lung carcinoma (NSCLC)	AMD3465	[99mTc]	(76)
Brevican (matrix protein) isoform dg-Bcan	Glioblastoma	BTP7	[18F]	(77)
Mutant K-Ras cells: K-Ras4B oncogenic isoform bind PDE-6	Colorectal cancer (CRC)	Inh of K-Ras4B/PDE-6 dissociation (I-C19)	[131I]	(78)
DLL3 Notch inhibitory ligand	Small cell lung carcinoma (SCLC)	Anti-DLL3 antibody SC16	[177Lu]	(79)
Carbonic anhydrase (CA9)4	CRC	Dual: CA9 inhibitor acetazolamide+CA9 probe	[111In]	(80)
	Renal cell carcinoma	CA9 Ab	[111In]	(81)
PDGFRβ	Carcinomas	Affibodies	[68Ga]	(82)
VPAC1 receptor	Epithelial cancers	VP2 small fusion protein binds VPAC1	[211At]5	(83)
	Bladder cancer	TP3805	[64Cu]	(84)
Tissue factor (TF)	Gastric cancers	Anti-TF Ab	[211At]	(85)
CD123	Acute leukaemia	Anti-CD123 Abs	[211At]	(86)
CD13+angiogenesis related amino-peptidase (APN)	Angiogenic tumour vessels	Dual: asparagine-glycine-arginine (NGR) +APD	[68Ga]	(87)
Death receptor 5 (DR5)	Colorectal cancer	Anti-DR5 Ab (CTB006)	[89Zr], [177Lu]	(88)
CDCP1	CRC	Anti-CDCP1 Ab	[89Zr]	(89)
Gastrin-releasing peptide receptor (GRPR)	Gastrointestinal stromal tumour (GIST)	NeoB–GRPR binding protein	[177Lu]	(90)

¹Fab, fragment antibody. ²Modest survival benefit for mice bearing HepG2 tumours but marked haematological toxicity due to long circulatory residence time of antibodies. ³Both internalised but greater uptake with whole Ab. ⁴Carbonic anhydrase is a cellular biomarker of hypoxia. ⁵High stomach uptake—late toxicity in mice.

Most targeted treatments are tailored to cancer type, but an approach that could lead to a pan-cancer TRT radiopharmaceutical exploits targeting the non-tumour cells present in the tumour microenvironment, particularly cancer-associated fibroblasts which overexpress the fibroblast-activating protein.

The tumour microenvironment, in addition to cancer cells, consists of immune cells, cancer-activated fibroblasts (CAFs), and host epithelial cells (91). Overwhelming evidence indicates that CAFs, or at least some CAF subtypes, are tumour promoting (92). Fibroblast activation protein (FAP) is a type II membrane-bound glycoprotein overexpressed in CAFs and is highly expressed in the stromal compartments of several malignant cancers (92).

The tumour inhibitory potential of FAP inhibitors has been explored but proved not particularly effective. However, FAP inhibitors radiolabelled with ⁶⁸Ga, e.g., [⁶⁸Ga]Ga-FAPI-46, are informative for the diagnosis and staging of several cancer types (93–97). Oster et al. (93) demonstrated in patients with glioblastomas a correlation between histological FAP expression and tumour SUV mean and SUV peak of [⁶⁸Ga]Ga-FAPI-46 using PET. Higher FAP expression was present in a gliosarcoma subgroup, suggesting that [⁶⁸Ga]Ga-FAPI-46 uptake may be useful diagnostically. Unterrainer et al. (94) demonstrated that [⁶⁸Ga]Ga-FAPI-46-PET could be beneficial in lymph node staging for patients with bladder cancer. Compared with CT alone, [⁶⁸Ga]Ga-FAPI-46-PET/CT improved the accuracy of staging resulting in major changes in the treatment of patients with oesophageal cancer (95) and pancreatic cancer (96). [⁶⁸Ga]Ga-FAPI-46-PET scans highlighted metastasis that were not evident in FDG-PET scans. However, for patients with NENs or prostate cancers, imaging of patients with [⁶⁸Ga]Ga-FAPI-PET is less informative than [⁶⁸Ga]Ga-DOTATATE (98) or [⁶⁸Ga]Ga-/[¹⁸F]F-PSMA-11 (99), respectively.

Recent studies have explored radiolabelled FAP inhibitors or antibodies to FAP for tumour treatment. Kuyumcu et al. (100) demonstrated that the mean absorbed dose to organs at risk with the ¹⁷⁷Lu-labelled quinoline-based inhibitor of FAP, FAPI-04, in four patients with metastatic cancers is reasonably low. FAPI-46, which has longer tumour residence times than FAPI-04 (101), radiolabelled with ⁹⁰Y (102–104) and ¹⁷⁷Lu (105), have been recently explored in dose-escalation studies with limited patient numbers. Nine patients with high FAP-expressing metastatic soft-tissue or bone sarcoma or pancreatic cancers who had exhausted other conventional treatments were given [⁹⁰Y]Y-FAPI-46 (102). Three patients received one treatment cycle, and six received two cycles. Radiographic disease control was evident in four patients. At-risk organ doses were low. In a larger study of a cohort of 21 patients with different cancers treated with [⁹⁰Y]Y-FAPI-46, partial response in one patient and stable disease in seven was reported (103). Thrombocytopenia and anaemia were evident in several patients. Patients with solitary fibrous tumours (11) received two or three cycles of [⁹⁰Y]Y-FAPI-46 (104). Disease control was evident in 9 of the 11 patients. A dose-escalation study of [¹⁷⁷Lu]Lu-FAPI-46 in 18 patients with FAP-expressing inoperable or refractory metastatic cancers (105) demonstrated good tolerance. Patients received 1.4 GBq increasing to 4.4 GBq of [¹⁷⁷Lu]Lu-FAPI-46 with no toxicity evident in most patients.

Short retention times compromise the therapeutic efficacy of some FAP inhibitors and potentially the efficacy of TRT via FAP. Several strategies to improve retention have been explored preclinically including ligand dimerization, covalent bond formation at the binding site, and the use of antibodies to FAP.

To improve binding, Zhong et al. (106) produced a dimerised version of FAPI-04 and compared tumour uptake of the monomer and dimer labelled with ⁶⁸Ga and ¹⁷⁷Lu using micro-PET. Images of SKOV3, A431, and H1299 xenografts revealed that tumour uptake of the dimer was greater than by the monomer. They also showed that [¹⁷⁷Lu]Lu(FAPI-04)(2) effectively reduced tumour growth. Pang et al. (107) compared tumour uptake of dimer and tetramer versions of FAPI-46 labelled with ⁶⁸Ga by FAP-expressing xenografts. They demonstrated far greater uptake and lower washout of the tetramer compared with the dimer. Tumour growth inhibition by [¹⁷⁷Lu]Lu-FAPI-04) was significantly (p<0.001) greater than by [¹⁷⁷Lu]Lu-FAPI-46 monomer. The first-in-human study of a [¹⁷⁷Lu]-labelled FAPI dimer was given to patients with treatment refractory breast, thyroid, or paraganglioma cancers (108). The study demonstrated increased residence time and higher median lesion absorbed dose using a dimer version of FAPi (6.7 Gy/GBq) compared with the monomer (0.6 Gy/GBq).

The FAPI FAP-2286 is a cyclic peptide with high-affinity FAP-binding characteristics. PET/CT scans of 21/21 patients with different malignancies including breast, pancreatic, and thyroid cancer demonstrated uptake of [⁶⁸Ga]Ga-FAP-2286 within primary solid tumours, adjacent excised tissues, and metastatic lesions (109), suggesting the potential of a general cancer tracer. The first-in-human study of [¹⁷⁷Lu]Lu-FAP-2286 for peptide-targeted radionuclide therapy (PTRT) was carried out on 11 patients with advanced adenocarcinomas of the pancreas, breast, rectum, or ovary with prior confirmation of uptake on [⁶⁸Ga]Ga-FAP-2286 or [⁶⁸Ga]Ga-FAPI-04 PET/CT. [¹⁷⁷Lu]Lu-FAP-2286 was well tolerated and demonstrated significant retained tumour uptake (110).

Most TRT is mediated through the interaction of a radionuclide conjugated ligand or antibody to an overexpressed cell surface receptor via a temporary non-covalent interaction. To increase the duration of target exposure to the radionuclide, radiopharmaceuticals that interact and form covalent bonds with the target receptor have been examined in preclinical work. Cui et al. (111) modified a FAP inhibitor by inclusion of an aryl fluorosulphate (FS) to covalently link with nucleophilic centres in the binding site of FAP. The FS-modified [¹⁷⁷Lu]Lu-FAPI demonstrated 2.5× higher tumour uptake, 13× longer retention, and significantly greater tumour control compared with [¹⁷⁷Lu] Lu-FAPI.

Xu et al. (112) developed a theragnostic pair targeting FAPα with two novel recombinant anti-FAPα antibodies labelled with ⁸⁹Zr and ¹⁷⁷Lu. PET/CT and SPECT/CT imaging of the ⁸⁹Zr and ¹⁷⁷Lu antibodies AMS002-1-Fc respectively demonstrated good tumour uptake by both and tumour control by [¹⁷⁷Lu] Lu-AMS002-1-Fc. Sibrotuzumab is an anti-FAP monoclonal antibody that was trialled for metastatic colorectal cancer but failed to show benefit. However, Xu et al. (112, 113) demonstrated using PET/CT, high uptake, and retention of the ⁸⁹Zr-labelled derivative of sibrotuzumab, PKU525, by FAP-expressing xenografts. They also showed correspondingly high tumour uptake of [¹⁷⁷Lu]Lu-DOTA-NCS-PKU525, e.g., 23% and 33% ID/g at 24 and 96 h, respectively, and achieved tumour growth inhibition with a single 3.7-MBq dose.

Liu et al. (114) produced a FAPI which could be radiolabelled with short t_1/2 imaging ¹⁸F, or therapeutic ²¹³Bi to pair the rapid kinetics. The inhibitor included an organotrifluoroborate linker which increased cell internalisation. FAPI has also been radiolabelled with ²¹¹At (115).

Early clinical trials of radiolabelled FAP ligands for the treatment of FAP-expressing cancers indicate that they are well tolerated. Developments in FAPI molecules that enhance tumour residence time may improve their therapeutic potential. Evidence suggests that prostate cancers and NENs are more effectively targeted, respectively, with PSMA ligands and somatostatin analogues than FAPIs. However, FAP targeting may be a useful way forward for the application of TRT to other cancer types.

Licensed cancer-targeting TRT radiopharmaceuticals

There are licensed TRT radiopharmaceuticals available to treat three types of malignancy, namely, B-cell lymphoma and neuroendocrine and prostate cancers. B-cell malignancies can be treated with licensed antibody-based TRT radiopharmaceuticals targeting upregulated receptors including CD20, whereas neuroendocrine and prostate cancers can be treated with licensed small-molecule TRT radiopharmaceuticals targeting the somatostatin receptor and PSMA, respectively.

Several anti-CD20 monoclonal antibodies such as rituximab, ofatumumab, or obinutuzumab improved the therapy of B-cell malignancies through mechanisms including cellular cytotoxicity and induction of apoptosis (116). When radiolabelled, these antibodies exploit the radiosensitivity of lymphoma cells with increased benefit (117). The first FDA-approved radiolabelled antibody was the [⁹⁰Y]Y-murine anti-CD20 antibody ibritumomab for the treatment of indolent lymphoma in 2002 (118), followed by [¹³¹I]I-tositumomab (Bexxar) in 2003 (119). Despite clear patient benefit including long-term remission in many patients and relatively low cost compared with other treatments for lymphoma, the uptake of these treatments is low due to a range of logistical reasons (119).

Most neuroendocrine cancers (NENs) overexpress the somatostatin receptor, which has several subtypes (120). The peptide ligand somatostatin is rapidly degraded in the circulation (121). Degradation-resistant somatostatin analogues including [¹¹¹In]In-DTPA-pentetreotide (Octreoscan^®) were developed in the 1980s and have been used for several decades for diagnosing neuroendocrine tumours (NENs) (122). However, image interpretation is complicated by high uptake by the liver, spleen, and kidneys. [¹¹¹In]In-DTPA-pentetreotide has been a treatment for NENs due to emission of Auger and internal conversion electrons by ¹¹¹In. This was replaced by β-emitter-labelled somatostatin analogues, e.g., [¹⁷⁷Lu]Lu-DOTA-TATE, under the name Lutathera^®, which can also be labelled with ⁶⁸Ga facilitating imaging by PET with improved resolution over SPECT (123). Positive outcomes reported in the NETTER trial for patients with midgut NENs treated with Lutathera^® led to approval by the FDA for SSTR2-positive gastroentero-pancreatic NETs (GEP-NETs) in adults (124).

Patients benefit from Lutathera^®, but many patients will relapse (125). Improving overall treatment outcome may be achieved by several routes. The [⁹⁰Y]-labelled version has been suggested for larger lesions as the β-emissions have a longer range than from [¹⁷⁷Lu] (126). Approaches using both [⁹⁰Y]- and [¹⁷⁷Lu]Lu-ligands may be advantageous (127), but combination studies are limited. Improving the radiopharmaceutical using antagonists ([¹⁷⁷Lu]Lu-DOTA-LM3 and -JR11) (128) which demonstrate greater receptor binding facilitates their application to cancers with lower SSTR density such as breast cancer (129). The use of radionuclides that produce more DNA damage such as the shorter range α-emitters including ²²⁵Ac may also improve outcomes and decrease normal tissue exposure (130).

Treatment of patients with prostate cancer is dependent on stage (131). Patients with advanced-stage metastatic castration-resistant prostate cancer (CRPC) may receive radionuclide treatments targeting prostate-specific membrane antigen (PSMA) (132). PSMA, a transmembrane protein, is also known as folate Hydrolase 1. As such, PSMA has a role in folate metabolism and internalises ligands. It is overexpressed on most prostate cancers and exploited as a target for the treatment of patients with CRPC (133). PSMA is commonly targeted with radiolabelled urea-based ligands particularly [¹⁷⁷Lu]Lu-PSMA-617 (134). Last-line salvage treatment of patients with bone involvement CRPC demonstrated that [¹⁷⁷Lu]Lu-PSMA-617 improves survival and is well tolerated (134). However, more than 50% of patients with CRPC do not show biochemical response (a 50% reduction in PSA) to [¹⁷⁷Lu]Lu-PSMA-617 (135). Refractory patients can receive PSMA-ligands labelled with α-emitter radiometals such as ²²⁵Ac (see section on α-emitters). Interestingly, a systematic review of several trials (136) identified that pretreatment with [¹⁷⁷Lu]Lu-PSMA-617 in some patients may induce resistance to treatment with [²²⁵Ac]Ac-PSMA-617.

Several normal tissues, including the salivary gland, also express PSMA attracting binding by radiolabelled PSMA ligands, which can result in loss of salivary function resulting in xerostomia (dry mouth) (137). This side effect is particularly frequent in patients receiving PSMA ligands labelled with α-emitters (137). Mitigation can be achieved by the co-administration of mono-sodium glutamate (MSG), but xerostoma is considered a dose-limiting toxicity for patients receiving radiolabelled PSMA-targeted ligands (137). A phase I clinical trial of [²²⁵Ac]Ac-labelled J591 reduces xerostomia and nephrotoxicity in metastatic castration-resistant PC (mCRPC) patients associated with radiolabelled PSMA-617 (138).

Strategies to improve the efficacy of TRT

Use of high linear energy transfer alpha emitters

Patients with prostate cancer where the disease has spread to bone, in the absence of visceral metastasis, may receive the bone-seeking agent radium-223 dichloride, [²²³Ra]RaCl₂, Xofigo^® (139). Radium is a calcium mimetic, and as such, ²²³Ra is concentrated in regions of bone turnover including adjacent to sites where metastatic cancers have established (139). Xofigo^®, the only α-emitter radiopharmaceutical to be clinically approved, is very effective at palliation of bone pain and has been shown to improve overall survival of patients with metastatic castration resistant prostate cancer (140).

Some patients with mCRPC can become resistant to ligands radiolabelled with β-emitters such as ¹⁷⁷Lu-PSMA but demonstrate response to α-emitters such as [²²⁵Ac]Ac-PSMA (18, 141). A phase I dose-escalation trial is currently underway to establish an optimal amount of activity for response that can be administered (141). Other [²²⁵Ac]-labelled therapeutics include [²²⁵Ac]Ac-DOTATOC and [²²⁵Ac]Ac-DOTA-substance-P, which have proven to be beneficial in patients with neuroendocrine tumours and gliomas, respectively (18).

Thorium-227 (²²⁷Th), which decays to ²²³Ra, can be chelated with octadentate 3,2-hydroxypyridinone (3,2-HOPO). Several conjugates of [²²⁷Th]Th-HOPO have been developed to target CD22-positive B cell cancers and CD33-positive leukaemia, and solid tumours overexpressing renal cell cancer antigen CD70 and membrane-anchored glycoprotein mesothelin in mesothelioma (142).

An α-decay event results in high-energy recoil (approximately 2% of the α-particle emission energy), which is sufficient to breach the integrity of the carrier radiopharmaceutical (143). Consequent release and free movement of the daughter species outside the tumour can result in irradiation of non-tumour tissue if the daughter undergoes further decays, sometime after the initial event, and contribute to toxicity (143).

²²⁵Ac (t_1/2 = 10 d, E = 6 MeV) produces six predominant radionuclide daughters in the decay cascade to stable bismuth-209 (²⁰⁹Bi) two high-energy gamma emissions, of which ²¹³Bi 440 keV is used for imaging (144). The ²²⁵Ac radionuclide daughters escape and circulate through the body and accumulate in different organs, and renal toxicity from released ²¹³Bi is the major concern (145). Some success with nanostructures to contain radionuclide daughters after α-emission has been achieved reducing release (146). Toro-González et al. (147) encapsulated [²²⁵Ac]Ac³⁺ chelated by a lipophilic 2,9-bis-lactam-1,10-phenanthroline ligand in poly(lactic-co-glycolic acid) (PLGA) nanoparticles, a biocompatible delivery platform used for drug delivery. Encapsulation within 155-nm PLGA nanoparticles was found to decrease the release of daughter species [²²¹Fr]Fr⁺ and [²¹³Bi]Bi³⁺ but only by approximately 50%. Karpov et al. (148) modified silica nanoparticles (SiO₂ NPs) with metallic shells composed of titanium dioxide (TiO₂) and gold (Au) nanostructures of 110 nm in size. In vivo and in vivo studies demonstrated that the metallic surface coating of SiO₂ NPs promotes an enhanced sequestering of radionuclides (²²⁵Ac and its daughter isotopes) compared with nonmodified SiO₂. However, clearance of these large nanoparticles is hepatobiliary, which is undesirable due to the radiosensitivity of the intestinal tract.

²¹¹At decays with the release of a single α-emission. Feng et al. (149) have demonstrated good tolerability of [²¹¹At]At-labelled PSMA ligands YF2 and L3-Lu in mice bearing xenografts, derived from prostate cancer cells. Laszlo et al. (86) developed anti-CD123+ targeting ²¹¹At-labelled antibodies, which extended the survival of mice bearing CD123+-expressing xenografts.

²¹²Pb initially undergoes β-decay to ²¹²Bi, which then decays by α-emission. The initial β-decay is accompanied by a high yield of conversion electrons and a cascade of Auger electrons converting the oxidation state of Bi to between Bi⁴⁺ to Bi⁷⁺, which has been shown to result in up to 40% release of daughter [²¹²Bi]Bi from chelation (150). However, preclinical work has demonstrated retention of a significant proportion of the disassociated [²¹²Bi] in the tumour environment (150).

Moving forward with radiopharmaceuticals carrying α-emitters including ²²⁵Ac and ²¹²Pb that release unchelated active daughters is likely to be challenging (151). Encapsulation in nanoparticles is an approach being explored to contain daughter species released by long-lived α-emitters, but due to the high recoil energy during α-decay, these need to be relatively large (~100 nm) (152). The large size of these particles results in hepatobiliary excretion and subsequent gut exposure (153). A further consideration with release of multiple daughter species is dosimetry. Encouragingly, a study has demonstrated that dual window planar imaging can discriminate ²²³Ra from ²²⁷Th in patients receiving ²²³Ra treatment (154), which enables accurate dosimetry to be determined. Clinical studies have demonstrated benefit and low toxicity of longer-lived α-emitters for example ²²⁵Ac to patients with advanced treatment refractory mCRPC (155). Exploration of α-emitting radionuclides towards the end of decay chains including ²¹²Bi and ²¹³Bi (t_1/2s ~1 h) have demonstrated, at least preclinically good tumour control. However, the therapeutic index from [²¹³Bi]Bi-PSMA-617 was reported to be lower than for [²²⁵Ac]Ac-PSMA-617 (156).

Pre-targeting approaches to radionuclide delivery

Antibody-targeted TRT radiopharmaceuticals are larger than the renal membrane and consequently have long blood clearance times, which can result in high levels of normal tissue radiation exposure (157). Pre-targeting is the initial administration of a complex consisting of an antibody conjugated, traditionally with avidin (part of an affinity pair with biotin) (see Figure 3 ). Following an interval of several days for tumour targeting and blood clearance of non-bound complex, the patient receives a radionuclide biotin conjugate, which is rapidly cleared from the circulation due to its small size (~1 kDa).

Figure 3.

Multistep pre-targeted procedure. *Avidin and **biotin have a strong mutual affinity.

Alternative affinity pairing systems include the Diels–Alder click reaction between tetrazine (Tz) and trans-cyclooctene (TCO) (158). The TCO-modified antibody binds with the radiolabelled Tz-substituted effector poly-L-Lysine. This system has been demonstrated to carry ²¹¹At for targeted alpha therapy and radioiodine for imaging. Poly-L-lysine was functionalised with a prosthetic group, for the attachment of both radiohalogens and tetrazine (159). As for direct targeting, pre-targeted approaches to TRT delivery require imaging using the same format of delivery system so that the imaging biomarkers accurately demonstrate tumour uptake of the therapeutic.

The theragnostic capability of the Tz/TCO pre-targeting system with the copper isotopes [⁶⁴Cu]- and [⁶⁷Cu]-radiolabelled Tz ([^64/67Cu]Cu-MeCOSar-Tz) was evaluated in a murine model of colorectal cancer (160). Mice pre-administered with the huA33 antibody modified with TCO (huA33-TCO) were treated 72 h later with [^64/67Cu]Cu-MeCOSar-Tz. Tumour uptake of [⁶⁴Cu] predicted response to treatment with [⁶⁷Cu]Cu-MeCOSar-Tz given either in a single dose or fractionated. Theragnostic approaches using the Tz/TCO system has also been examined using a self-assembling and disassembling molecular system for treating glioblastoma using murine models (161).

Combination of TRT with immune activation and radiosensitisation

The primary mechanism of cell death induced by ²²³Ra is by induction of DNA damage particularly double-strand breaks (DSBs) (162). Other mechanisms that may contribute to the therapeutic efficacy of Xofigo^® include immunogenic mechanisms (163). DNA damage can activate the stimulator of interferon gene (STING) signalling pathway activating NLRP3-dependent pyroptosis, a typical form of immunogenic cell death, to enhance antitumor immune response. This may be an important mechanism of tumour control induced by ²²³Ra-irradiation (163). Biomarkers that can guide selection of patients for treatment with PD-1/PD-L1 inhibitors include PD-1/PD-L1 expression and mismatch repair deficiency (164).

TRT treatments may enhance immune response through enhancement of infiltration of CD4⁺ and CD8⁺ T cells (165). Zboralski et al. (166) have shown that combined treatment of mice bearing FAP-expressing xenografts with a [¹⁷⁷lu]Lu-labelled FAP inhibitor ([¹⁷⁷Lu]Lu-FAP-2287) and a PD-1 inhibitor increased recruitment of tumour-infiltrating CD8(+) T cells.

Clinical trials combining immune checkpoint inhibitors (ICIs) with TRT are underway including the STARLITE 2 Phase 2 trial of patients with clear cell renal carcinoma treated with the ICI, nivolumab, and [¹⁷⁷Lu]Lu-girentuximab, which is an anti-carbonic anhydrase IX antibody (167).

Interestingly, the type of radionuclide influences the synergistic therapeutic effect of TRT with PD-1 inhibitors. Compared with monotherapy, combination treatments of PD-1 inhibitors with [²¹³Bi]-anti-melanin and to a lesser extent [¹⁷⁷lu]Lu-anti-melanin, but not [²²⁵Ac]-anti-melanin, increased the treatment response of melanomas (168).

Chimeric antigen receptor T cells (CART) are T lymphocytes that have been taken from a patient and reprogrammed to identify and attack their cancers (169). CARTs are relatively ineffective against solid tumours, but this can be increased by exposure of the cancers to radiation (170). Sodji et al. (170) compared the effector (cytotoxic activity against GD2 expressing human neuroblastoma (CHLA-20) and M21 melanoma cells) and viability of anti-GD2 CART cells after exposure to different doses from ²²⁵Ac or ¹⁷⁷Lu. Radiation enhanced the cytotoxic activity of these CAR T cells against CHLA-20 and M21 independent of dose tested and type of radionuclide. ²²⁵Ac was more toxic than ¹⁷⁷Lu to anti-GD2 CAR T cells, suggesting that ¹⁷⁷Lu-based TRT may be preferred over ²²⁵Ac-based TRT to enhance CAR T activity.

Radiosensitisers are commonly used to improve the therapeutic efficacy of external beam radiotherapy (171). Radio-sensitisation can be mediated through many mechanisms including chemotherapy and inhibition of DNA damage response proteins such as ATM and DNA-PKcs and DNA repair enzymes, e.g., topoisomerase I and PARP inhibitors. Phase 1 trials of radiotherapy with the potent ATM/DNA-PKcs inhibitor XRD-0394 has recently reported favourable results (172). Beneficial patient outcomes of radiosensitisers with radiation therapy are a strong rationale for combination treatments with molecular radiotherapy.

Antimetabolite chemotherapy drugs including 5-fluorouracil (5FU) and capecitabine are commonly used in combination with radiotherapy (171). Combination treatments of patients with neuroendocrine tumours with [¹⁷⁷Lu]Lu-DOTATATE and 5FU or capecitabine are well tolerated (173). Some studies have demonstrated that combining TRT with capecitabine improves patient outcome (174).

Response to radiation-based treatments is countered by DNA repair mechanisms (175). Targeting DNA repair enzymes is an established form of radiosensitisation (176). Poly(ADP-ribose) polymerase 1 (PARP-1) is essential in DNA single-strand break (SSB) repair (see Figure 4 ). Several PARP inhibitors (PARPi) are now clinically approved (176). These bind to PARP/SSB complexes formed on damaged DNA preventing repair and disengagement of the repair complex, which in turn results in double-strand break (DSB) formation. BRCA1/2 mutant cancers have a limited capacity to repair DSBs. PARP inhibitors are particularly effective for patients with mutant BRCA1/2 cancers (177). Combination treatment of a murine model of a triple-negative breast cancer with the radiolabelled FAP inhibitor [¹⁷⁷Lu]Lu-DOTAGA.(SA.FAPi)(2) and PARP inhibitor olaparib increased therapeutic efficacy over [¹⁷⁷Lu]Lu-DOTAGA.(SA.FAPi)(2) alone (178). Clinical trials (early phase) assessing combination PARPi + [177Lu]Lu-DOTATATE for pancreatic and metastatic NENs are ongoing (173).

Figure 4.

Mode of action of poly ADP-ribose and its inhibition during repair of single-strand DNA breaks.

Tyrosine kinases (TK) are components of receptors and cell signalling pathways, many of which induce cell growth and proliferation. Anticancer treatments include TK inhibitors (TKI) (173). Some TKIs are multi-TK inhibitor, such as sunitinib, and can inhibit angiogenesis as well as cancer cell growth. Early phase trials of combination treatments of sunitinib and other TK inhibitors with NEN-targeted TRT are ongoing (173).

To discover potential strategies for sensitisation to TRT, Qin et al. (179) examined protein activation in cholecystokinin B receptor expressing A431 cancer cells treated with a ²²⁵Ac-labeled minigastrin analogue ([²²⁵Ac]Ac-PP-F11N) using proteomics and phospho-proteomics. They identified several upregulated proteins associated with carcinogenesis and DNA repair pathways including histone deacetylases (HDAC) for which there is an FDA-approved inhibitor, SAHA. Treatment of A431/CCKB cells with SAHA and [²²⁵Ac]Ac-PP-F11N was synergistic over treatment with each alone delivering increased DSB formation and tumour cell kill.

Many studies have demonstrated that hyperthermia can improve response to radiotherapy (180). Heating is delivered through several techniques including ultrasound to raise local temperature to 40 °C-43°C. Mechanisms responsible for the beneficial effect of hyperthermia include improved blood flow, inhibition of DNA repair, and immune activation (176). Gold nanoparticles have the property of near IF light (NIR) to heat conversion, known as the photothermal effect (PT). When incorporated into tumour tissue and activated by NIR, AuNPs can achieve localised and targeted hyperthermia (181). Simón et al. (181) demonstrated that PT, delivered with NIR irradiation of xenografts in mice administered with AuNPs, improved the therapeutic efficacy of [¹⁷⁷Lu]Lu-DOTA-TATE PRRT against SSTR-expressing cancers in mice.

DNA irradiated with ionising radiation is damaged either by direct interaction with radiation or via the formation of free radicals particularly reactive oxygen species (ROS) (182). Photodynamic therapy (PDT) is mediated by the formation of ROS from molecular oxygen by the transfer of energy from photosensitisers excited by light at specific wavelengths (183). Generated ROS will then contribute to DNA damage. A precursor of the photosensitiser, protoporphyrin IX (PpIX), 5-aminolevulinic acid (5-ALA), accumulates in cancer tissue (184). Jo et al. (184) demonstrated that photodynamic therapy can be elicited using Cerenkov luminescence energy transfer (CLET) from the decay of ⁶⁴Cu to PpIX in mice administered with [⁶⁴Cu]Cu-DOTA-trastuzumab and 5-ALA for high-precision PDT of HER-2-overexpressing cancer.

Gold nanoparticles can enhance radiosensitivity (185). Cysteine functionalised glutathione-coated gold nanoparticles, l-Cys-GSH-AuNP, aggregate in the acidic conditions in lysosomes facilitating lysosomal accumulation on internalisation by cancer cells. PET imaging of xenograft-bearing mice injected with [⁶⁸Ga]Ga-l-Cys-GSH-AuNP demonstrated long-term tumour accumulation and increased response to external beam radiotherapy (185).

Discussion

TRT is a relatively recent addition to the anticancer treatment repertoire. In the last few years, the phase 3 NETTER and VISION trials of [¹⁷⁷Lu]-radiolabelled TRT agents for neuroendocrine cancer and prostate cancer respectively have reported beneficial outcomes in terms of improved progression free survival (PFS) (NETTER) (124) and overall survival and PFS (VISION) (162). These findings have greatly increased interest in TRT especially for prostate cancer treatment and expansion to other cancer types. This has brought into focus the requirement for reliable supplies of medical radionuclides and the need for an expansion in production facilities.

Pretreatment imaging of candidate patients is essential to demonstrate suitability for TRT and inform on dosimetry (48). Ideally, imaging should be carried out with an imaging component of a theragnostic pair to provide biodistribution and dosimetry information that will most accurately predict the pharmacokinetics of the therapeutic (49).

Combining TRT with other treatments can improve efficacy increasing long-term benefit or allow for lower radioactive doses reducing toxicity and helping to minimise the number of patients who withdraw from treatment. Radiosensitisers are commonly used with external beam radiotherapy (EBRT) to improve response but could be applied alongside TRT (186). Predictive biomarkers can be used clinically to inform on radiosensitiser use for patients with hypoxic, and therefore radioresistant, cancers (187).

Approximately 25% of patients with lymphoma treated with [⁹⁰Y]Y- 90 Y-ibritumomab tiuxetan (⁹⁰Y-IT, Zevalin, Acrotech Biopharma) demonstrate complete long-term remission (119). Many patients with mCRPC patients benefit from [¹⁷⁷Lu]PSMA-617, but approximately 30% do not respond. These findings illustrate the need for predictive biomarkers based on intrinsic tumour biology to stratify patients for treatment modalities. Several cancer types have been shown to be differentiable into molecular subtypes (188). The molecular subtype is associated with response or resistance to different treatment types; for example, some subtypes of bladder cancer are resistant to chemotherapy but sensitive to other treatments (188). Studies that stratify TRT response by molecular subtype may help to define which patients may benefit from TRT or require sensitisation. Other approaches include the use of intrinsic radiosensitivity gene expression signatures as developed for external beam radiotherapy (189). Studies have examined TRT-upregulated genes which may help to identify candidate signature genes (190). These approaches will require multiple large patient cohorts for validation. Applying biomarkers to combination treatments will require an assessment of suitability of the patient for each treatment type using multiple biomarkers.

Some patients, refractory to β-emitter-radiolabelled PSMA and somatostatin ligands, have demonstrated beneficial response to ligands radiolabelled with α-emitters such as ²²⁵Ac (125). Daughter decay products from the longer-lived α-emitters like ²²⁵Ac tend to be released from their chelate and associated with normal tissue toxicity. Fortunately, toxicity associated with [²²⁵Ac] Ac-PSMA-617 treatment of patients with mCRPC is considered low (155). However, supply of ²²⁵Ac is a serious issue. A further limitation in some European countries is the requirement for hospital confinement for several days post-treatment reducing the number of patients who can be treated at any one time.

Inhomogeneous dose distribution is a common problem with TRT due to the heterogeneity in target expression between cancer cells, even within the same tumour, and regions of poor blood perfusion due to the rudimentary nature of the tumour vasculature (191). Dose distribution inhomogeneity may result in undertreatment of some cancer cells and inevitable tumour recurrence. The crossfire effect, which is the deposition of energy emitted from a radionuclide into distant cells, particularly from β-emitters, helps to even out dose across tumours. Combinations of β-emitters with different energies will result in a greater spread of dose and may further help in overcoming dose inhomogeneities (127).

Hypoxic cancer cells which are resistant to low LET radiation but sensitive to high LET radiation tend to be distributed as single cells or small groups of cells (192). Alpha-particles with their short range and highly damaging characteristics within a few cell diameters are ideal for targeting hypoxic cells in tumours.

TRT is an exciting approach to cancer treatment with many variation possibilities including use of multiple radionuclides to improve dose distribution, combined with other treatments to synergise treatment efficacy to bring about its optimisation. Validated biomarkers to identify patients most likely to benefit or conversely who may not benefit from TRT are crucial. The capacity to produce and deliver medically radionuclides particularly, ¹⁷⁷Lu, for treatment is essential to ensure that these effective treatments can be expanded to fulfil the predicted demand.

Funding Statement

The author(s) declare that no financial support was received for the research, and/or publication of this article.

Author contributions

TS: Writing – original draft, Writing – review & editing.

Conflict of interest

The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.