Prostate-specific membrane antigen as a target for cancer imaging and therapy

Abstract

The prostate-specific membrane antigen (PSMA) is a molecular target whose use has resulted in some of the most productive work toward imaging and treating prostate cancer over the past two decades. A wide variety of imaging agents extending from intact antibodies to low-molecular-weight compounds permeate the literature. In parallel there is a rapidly expanding pool of antibody-drug conjugates, radiopharmaceutical therapeutics, small-molecule drug conjugates, theranostics and nanomedicines targeting PSMA. Such productivity is motivated by the abundant expression of PSMA on the surface of prostate cancer cells and within the neovasculature of other solid tumors, with limited expression in most normal tissues. Animating the field is a variety of small-molecule scaffolds upon which the radionuclides, drugs, MR-detectable species and nanoparticles can be placed with relative ease. Among those, the urea-based agents have been most extensively leveraged, with expanding clinical use for detection and more recently for radiopharmaceutical therapy of prostate cancer, with surprisingly little toxicity. PSMA imaging of other cancers is also appearing in the clinical literature, and may overtake FDG for certain indications. Targeting PSMA may provide a viable alternative or first-line approach to managing prostate and other cancers.

Accurate diagnosis, staging, and therapeutic monitoring in patients with cancer are the most important goals of oncologic imaging. Those goals are often well served with functional/metabolic imaging, particularly within the realm of nuclear medicine with positron-emission tomography (PET) and single-photon emission computed tomography (SPECT), leveraging a variety of radiolabeled compounds. PET imaging with 2-deoxy-2-[¹⁸F]fluoroglucose (FDG) has proven to be transformative in oncologic imaging in recent decades.^1–4 However, FDG PET has profound limitations in imaging certain malignancies including prostate cancer,^5–7 spurring the need for development of new radiotracers able to address these limitations. That is illustrated, in part, by the emergence of numerous PET and SPECT radiotracers targeting the prostate-specific membrane antigen (PSMA).

Conventional imaging of primary prostate cancer is generally performed with magnetic resonance (MR) imaging, with conventional imaging of metastatic or recurrent prostate cancer centered around contrast-enhanced computed tomography (CT) and bone scan with ^99mTc-methylene diphosphonate (MDP). MR imaging can be non-specific in the context of primary prostate cancer, with benign prostatic hyperplasia at times having similar imaging characteristics to malignant tissue.⁸ Both CT and bone scan have limitations in sensitivity and specificity for the detection of early/subtle recurrence or metastasis, e.g., sub-centimeter, disease-involved lymph nodes and sclerotic bone metastases. High-affinity radiotracers targeting PSMA could potentially address those limitations.

PSMA represents an excellent target for molecular imaging of prostate cancer. PSMA is a type II membrane metalloenzyme that exhibits developmentally controlled and tissue-specific expression patterns (Figure 1).⁹ Expression on the plasma membrane is restricted to a few healthy tissues such as lacrimal and salivary glands, proximal renal tubules, epididymis, ovary, the luminal side of the ileum-jejunum and astrocytes within the central nervous system (CNS); healthy prostate gland expresses comparatively little PSMA, which is confined within the apical epithelium of secretory ducts.^10–12 In these non-malignant tissues, uptake of PSMA-targeted probes may be limited by an intact blood-brain barrier, a healthy proximal small bowel lumen, and truncated cytoplasmic expression of PSMA within normal prostate. PSMA within prostate cancer cells begins to up-regulate and migrate to the plasma membrane during the transition to androgen independence, and is most associated with high grade, metastatic disease.^13–16 Nevertheless, PSMA is expressed in most primary prostate tumors as well, regardless of androgen status.^17,18

Figure 1.

Homodimer of human PSMA (crystal structure) tethered to the biological membrane. One monomer shown in semitransparent surface representation with individual domains of the extracellular part colored green (protease domain; amino acids 57 – 116 and 352 – 590), blue (apical domain; amino acids 117 – 351), and yellow (C-terminal; amino acids 591 – 750); the second monomer is colored gray. N-linked sugar moieties are colored cyan, and the active-site Zn²⁺ ions are shown as red spheres. Left panel: residing at the plasma membrane of astrocytes /schwann cells, PSMA catabolizes N-acetylaspartyl glutamate (NAAG), the most prevalent peptidic neurotransmitter in the mammalian nervous system. N-acetylaspartate and glutamate, the reaction products, are selectively transported into glial cells, metabolized and reused for NAAG synthesis in neurons. Right panel: PSMA (or folate hydrolase) at the plasma membrane of enterocytes in the proximal jejunum sequentially hydrolyzes the C-terminal γ-glutamate tail of dietary folates, finally leaving folate-monoglutamate, which can then be transported transcellularly into the blood stream [Adapted from Barinka C et al.].⁹

Because the active site of PSMA is highly conserved,¹⁹ the development of molecular probes binding with high affinity and specificity to the active site is an efficient strategy that avoids dependence on glycosylation patterns^20–22 and other post-translational, cell-specific processing, which may be subject to the tumor microenvironment. The caveolin-dependent, rapid internalization of PSMA while bound as a dimer to its ligand²³ is also a desirable feature of this target, as well as its final peri-nuclear localization.^24,25 Endogenous substrates include dietary poly-γ-glutamyl folates ^26,27 and N-acetylaspartyl glutamate (NAAG, within CNS),^28,29 but the function and identity of other ligands is a topic of speculation. PSMA is also expressed within the neovascular endothelium of most solid tumors,^30–34 enabling imaging of a variety of malignancies with probes targeted toward PSMA.

This review will provide a detailed account of the structure-based design of small-molecule PSMA inhibitors, including radiohalogenated agents and radiometals for radionuclide imaging, as well as other PSMA-targeted agents for optical and MR imaging. It will then describe the clinical experience to date with PSMA-targeted radionuclide imaging and the emergence of potential PSMA-targeted therapies and theranostic agents.

Structure-based design of small-molecule PSMA inhibitors

The high expression of PSMA (EC 3.4.17.21), a.k.a., glutamate carboxypeptidase II (GCPII), N-acetyl-L-aspartyl-L-glutamate peptidase I or NAAG peptidase, on the surface of prostate cancer cells and on most tumor neovasculature has rendered it a useful and popular target for molecular imaging. Small-molecule PSMA inhibitors are zinc binding compounds attached to a glutamate or glutamate isostere and fall into three families: 1) phosphonate-, phosphate-, and phosphoramidate compounds; 2) thiols; and 3) ureas.^35,36 Initial phosphonate and phosphate inhibitors were prepared at ZENECA and later at Guilford Pharmaceuticals,^37,38 and phosphoramidate compounds were reported by Berkman et al.^39–41 Urea-based compounds designed to inhibit GCPII in the brain were first reported by Kozikowski et al.^42,43 Low-molecular-weight agents for imaging and therapy of PSMA prepared to date have been either ureas or phosphoramidates, with the majority having been based on the urea scaffold. Those agents possess a terminal glutamate or glutamate isostere at the P1′ position, which enables productive binding to PSMA, and are amenable to modification with bulky substituents that interact with the arginine patch or tunnel region of PSMA.

Recent crystallographic studies of complexes between PSMA and low-molecular-weight ligands elucidated their binding modes within the active site of PSMA, revealing that the S1 pocket of PSMA is more tolerant toward structural modification than the glutamate-binding S1′ (pharmacophore) pocket.⁴⁴ There is also a deep, funnel-shaped accessory tunnel that contains several more exosites that could be exploited in design of more potent inhibitors. Barinka et al. reported the first detailed structures of complexes between human PSMA and urea-based inhibitors and identified a hydrophobic accessory pocket near the S1 site.⁴⁵ That pocket produced an unusually high binding interaction with 2-[3-[1-carboxy-5-(4-iodo-benzoylamino)-pentyl]-ureido]-pentanedioic acid (DCIBzL) (Table I, Entry 5), one of the most potent urea-based inhibitors of PSMA synthesized to date (K_i=0.01 nM).⁴⁶ The crystal structure of PSMA complexed with DCIBzL (PDB ID: 3D7H) revealed that the iodophenyl group of DCIBzL projects into the arginine patch region and resides in the accessory pocket created by side chains of arginines 463, 534 and 536 (Figure 2).⁴⁵ The hydrophobic pocket accessory to the S1 site has recently been exploited for structure-based design of new PSMA inhibitors to improve binding affinity and pharmacokinetic properties.^47–49

Table I.

Carbon-11 and Radiohalogen-based, PSMA-targeted PET and SPECT imaging agents.

Entry	Structure	Ref.	Entry	Structure	Ref.
1		51	2		52
3		53	4		46
5		46	6		46
7		68	8		69
9		49	10		64
11		54	12		55
13		55	14		55
15		57	16		70
17		59	18		66
19		61	20		62
21		171	22		172
23		59	24		60
25		56	26		56
27		65

Figure 2.

The hydrophobic pocket accessory to the S1 site. The active site bound DCIBzL is in stick representation. The dissected substrate-binding cavity of PSMA is shown in semi-transparent surface representation (gray). The side chains of amino acids delineating the “accessory hydrophobic pocket” are shown in stick representation and colored cyan. The active-site Zn⁺² and S1 bound Cl⁻ are colored blue and represented as a transparent sphere, respectively and water molecules are shown as red spheres. Adapted from Barinka C et al.⁴⁵

In another approach, Zhang et al. identified and structurally characterized another exosite of PSMA that binds aromatic moieties.⁵⁰ That exosite, termed the arene-binding site, is formed by the indole group of Trp541 and the guanidinium group of Arg511. Attaching a dinitrophenyl moiety with a length-optimized linker to a PSMA inhibitor significantly enhanced affinity toward PSMA through the avidity effect of the arene-binding site, namely, by allowing it to bind to PSMA in a bi-dentate mode by interacting with both S1′ and S1 pockets.

Radiolabeled small-molecule PSMA inhibitors for radionuclide imaging

We have divided this topic into two sections, one focusing on radiohalogenated agents and the other on those employing radiometals, rather than by modality. That reflects the possibility of a particular scaffold being used for more than one modality.

Radiohalogenated agents

A list of radiohalogenated and ¹¹C-labeled small-molecule PSMA inhibitors is presented in Table I. The first reported radiolabeled small-molecule PSMA inhibitor for PET imaging was the methyl cysteine-glutamate urea, [¹¹C]MCG, a.k.a., N-[N-[(S)-1,3-dicarboxypropyl]carbamoyl]-S-[¹¹C]methyl-L-cysteine, or [¹¹C]DCMC, (Table I, Entry 1), which demonstrated PSMA-specific uptake in mouse and non-human primate kidney ⁵¹ and in PSMA+ LNCaP tumor xenografts.⁵² Foss and coworkers demonstrated that a radiohalogenated small-molecule PSMA inhibitor, the iodotyrosine-glutamate urea N-[N-[(S)-1,3-dicarboxypropyl]carbamoyl]-S-3-[¹²⁵I]iodo-L-tyrosine, [¹²⁵I]DCIT (Table I, Entry 2), also showed PSMA-specific uptake in PSMA+ LNCaP tumor xenografts.⁵² Expanding to ¹⁸F, that same group synthesized N-[N-[(S)-1,3-dicarboxypropyl]carbamoyl]-4-[¹⁸F]fluorobenzyl-L-cysteine ([¹⁸F]DCFBC) (Table I, Entry 3) to leverage the salutary characteristics of this isotope for PET and clinical translation.⁵³ [¹⁸F] DCFBC enabled visualization of PSMA+ PC3 PIP tumor xenografts with low normal organ uptake except for kidney and accumulation within bladder. A second generation agent, 2-(3-(1-carboxy-5-[(6-[¹⁸F] fluoro-pyridine-3-carbonyl)-amino]-pentyl)-ureido)-pentanedioic acid [¹⁸F]DCFPyL (Table I, Entry 11), which contains a 6-[¹⁸F]fluoronicotinamido group on a lysine-glutamate urea, gave higher tumor to normal tissue ratios compared to [¹⁸F]DCFBC in PSMA+ PC3 PIP tumor xenografts (Figure 3).⁵⁴ Other ¹⁸F-labeled, aryl substituted lysine glutamate ureas have been prepared (Table I, Entries 4, 12, 13, 14, 25, 26).^46,55,56 Substituted Glu-Glu ureas have also been radiolabeled with ¹⁸F. Those include a 6-[¹⁸F] fluoronicotinamido substituted linker-Glu-Glu urea (Table I, Entry 15),⁵⁷ Al¹⁸F-2,2′,2″-(1,4,7-triazonane-1,4,7-triyl) triacetic acid (NOTA) linker-Glu-Glu ureas (Table I, Entries 17 and 23),^58,59 and Al¹⁸F-2-[2-[carboxymethyl-[(2-hydroxyphenyl)methyl] amino]ethyl-[(2-hydroxyphenyl)methyl]amino]acetic acid (HBED)-linker-Glu-Glu urea (Table I, Entry 24).^{60 18}F-Labeled Phe-Glu ureas (Table I, Entries 19 and 20) have also been prepared.^61,62 Compound P238 (Table I, Entry 20) is dimeric with a high affinity for PSMA (K_i=0.1–0.4 nM), and demonstrated high specific uptake in PSMA+ LNCaP tumor xenografts with fast clearance from normal tissues through the kidneys.⁶³ Two ¹⁸F-labeled phosphoramidates (Table I, Entries 10 and 27) have been reported.^64,65 Both agents demonstrated specific uptake in PSMA expressing tumor xenografts with fast clearance from normal tissues. An ¹⁸F-labeled derivative (Table I, Entry 18) of the potent PSMA inhibitor 2-(phosphonomethyl)pentandioic acid (2-PMPA) has also been reported.⁶⁶ Although that agent clearly visualized PSMA+ LNCaP tumor xenografts, the images also showed significant bone uptake likely due its phosphonate structure.⁶⁷ Several radioiodinated Lys-Glu urea derivatives have been reported that utilized different radioiodinated prosthetic groups. Those include the 4-iodo-benzoyl group (Table I, Entry 5),⁴⁶ the 5-iodo-3-carbonyl-pyridine group (Table I, Entry 6),⁴⁶ the 4-iodo-benzyl group (Table I, Entry 7),^{68, 69} the 4-iodo-phenyl-ureido group (Table I, Entry 8) ^{68, 69} and an iodotriazole (Table I, Entry 16).⁷⁰ All of those radioiodinated agents demonstrated specific uptake in PSMA expressing tumor xenografts and roughly similar normal organ biodistribution with the exception of the iodotriazole compound (Table I, Entry 16), which had significant uptake in the thyroid (10% of the injected dose per gram of tissue [% ID/g] at 23 hours postinjection). Compounds MIP-1072 Table I, Entry 7) and MIP-1095 (Table I, Entry 8) have also been radiolabeled with ¹²³I and translated to clinical imaging where they enabled detection of metastatic prostate cancer with high sensitivity.⁷¹ MIP-1095 has also been prepared with ¹²⁴I for PET imaging and with ¹³¹I for radiopharmaceutical therapy, with clinical trials ongoing, as discussed below.⁷²

Figure 3.

PET/CT volume-rendered composite images of [¹⁸F]DCFPyL in a mouse bearing PSMA+ PC3 PIP and PSMA− PC3 flu tumors at (A) 0–30 minutes; (B) 30–60 minutes and (C) 3–3.5 hours post-injection. By 30 minutes radiochemical uptake was evident within the PSMA+ PC3 PIP tumor and kidneys. Radioactivity receded from kidneys faster than from tumor, and was not evident within kidneys by 3.5 hours post-injection. Radioactivity within bladder was due to excretion. At no time was radioactivity clearly visualized within the isogenic, PSMA-PC3 flu tumor. Adapted from Chen Y et al.]⁵⁴

Radiometals

Several classes of PSMA-based, radiometal-labeled imaging agents using Lys-Glu-urea as the targeting scaffold have been reported. To provide high-affinity agents, Banerjee et al., envisioned the need to attach a linker between a bulky metal chelator and the PSMA-targeting urea moiety, allowing the urea to reach the binding site while keeping the bulky metal chelated part on the exterior of the enzyme active site (Figure 4). By employing that targeting strategy they have synthesized a series of SPECT imaging agents using both ⁹⁹mTc^I(CO)₃ and ^99mTc^VO⁺³ labeling techniques and several well-studied chelating agents related to the individual coordination chemistry required for these two species (Table II, Entries 2–4, 11–12).^{73, 74} Although most of those agents demonstrated high uptake and retention in PSMA expressing xenografts, the choice of chelating agent was found to have a moderate to significant effect on the overall pharmacokinetics. Low et al. reported a series of ^99mTc-Oxo-labeled agents, the lead agent (Table II, Entry 5) having been prepared using bis-Glu urea as the targeting moiety.⁷⁵ Those compounds showed specific uptake in PSMA expressing xenografts. Babich et al. investigated a series of ^99mTc and Re-tricarbonyl-based agents using both Lys-Glu as well as Glu-Glu ureas as targeting scaffolds with different chelating agents having different degrees of hydrophilicity (Table II, Entries 7, 9–10).^76–78 Agent ^99mTcMIP-1404 exhibited the best combination of high tumor uptake and rapid clearance from kidney and non-target tissues by four hours post-injection.^{77, 78} Phosphoramidate-based inhibitors were also investigated using the ^99mTc-tricarbonyl-core and 2-[bis[2-[bis(carboxymethyl)amino]ethyl]amino]acetic acid (DTPA) and bis(pyridin-2-ylmethyl)amine (DPA) as chelating agents (Table II, Entries 6 and 8).^{79, 80} Misra and coworkers reported a simple solid phase, cartridge-based method to prepare ^99mTc- S-acetylmercaptoacetyltriserine-NHS from ^99mTcO₄⁻ to ~99% radiochemical purity.⁸¹ The radiolabeled intermediate was then conjugated to negatively charged adamantane-trimerized 2[(3-ami-no-3-carboxypropyl)(hydroxy)(phosphinyl)-methyl] pentane-l,5-dioic acid (GPI) in one step without the need for purification by high performance liquid chromatography to produce ^99mTc-labeled agents (Table II, Entry 1). PSMA-specific cell binding was verified but no in vivo data are available.

Figure 4.

Long linker strategy for attachment of radiometals and other bulky substituents to PSMA-targeted, small-molecule agents. Tri-partite construct including the chelator, linker and targeting moiety (A). A minimal linker length of approximately 20Å enables productive binding of the construct to PSMA through externalization of the bulky substituent (B).

Table II.

Radiometal-based, PSMA-targeted PET and SPECT imaging agents.

Entry	Structure	References
1		81
2		73
3		73
4		73
5		75
6		79
7		76
8		80
9		77
10		78, 124
11		74
12		74
13		92
14		92

Using the linker-urea PSMA-targeting construct, Banerjee et al, have extended their work to explore generator-produced ⁶⁸Ga for PET imaging and reported the synthesis and testing of the ⁶⁸Ga-labeled PSMA-targeted imaging agents depicted in Table III, Entries 1–2.⁸² They chose the commercially available macrocyclic chelating agent DOTA-monoamide so that the agents could be used for both imaging and therapeutic applications. The lead agent from that initial study, [⁶⁸Ga]6 (Entry 2), demonstrated uptake in PSMA+ PC3 PIP tumors at 45 minutes post-injection with little visible uptake in PSMA-PC3 flu xenografts.⁸² That agent also showed fast renal clearance while maintaining high PSMA+ tumor retention with tumor to muscle and tumor to blood ratios of 110 and 22, respectively. Eder et al. reported [⁶⁸Ga]DKFZ-PSMA-11, i.e. ⁶⁸Ga-PSMA, Table III, Entry 9),⁸³ using the chelator N,N′-bis[2-hydroxy-5-(carboxyethyl)-benzyl]ethylenediamine-N,N′-diacetic acid (HBED-CC), an analog of HBED.^84–87 HBED is a potentially more attractive chelator for ⁶⁸Ga because it forms a more thermodynamically stable complex than does DOTA (logK_ML of 35.6 vs. 21.3 for DOTA) ⁸⁷. Furthermore, DKFZ-PSMA-11 can be radiolabeled with ⁶⁸Ga(III) at room temperature in less than five minutes in high yield and purity. When similar post-radiolabeling purification schemes for [⁶⁸Ga]6 (Table III, Entry 2) ⁸² and [⁶⁸Ga]DKFZ-PSMA-11 were employed, the two compounds displayed similar uptake and retention in PSMA+ PC3 PIP tumors, although significantly higher tumor to kidney, tumor to spleen, and tumor to salivary gland ratios were observed for [⁶⁸Ga]6.⁸⁹ A recent report by Eder and coworkers (Table III, Entry 15) also showed rapid renal clearance and normal tissue uptake of the ⁶⁸Ga-DOTA-mono amide analog [⁶⁸Ga]PSMA-6l7 compared to [⁶⁸Ga]DKFZ-PSMA-11, which bears the HBED-CC chelator.⁹⁰ That group has also used the HBED-CC chelator in a ⁶⁸Ga-labeled hetero-bivalent compound designed to target both PSMA and gastrin-releasing peptide, and have demonstrated dual-targeted PET imaging in pre-clinical models.⁹¹

Table III.

Radiometal-based, PSMA-targeted theranostics

Compound	Entry	References
	1	82
	2–4 5	82, 92, 175 153, 154
	6–8	158, 108
	9	83
	11, 12	153, 154
	13 14	155 154
	15	89

Banerjee et al. have also investigated the cyclotron-produced, long-lived (T_1/2=12.7 hours) radionuclide ⁶⁴Cu for PET imaging of PSMA ⁹² by utilizing different macrocyclic chelating agents as shown in Table II, Entries 13 and 14. The agent conjugated with the 4,8,11-tetraazabicyclo[6.6.2]hexadecane-4,11-diacetic acid (CB-TE2A) chelator (Table II, Entry 14), demonstrated improved biodistribution with rapid clearance from normal tissues, including kidney, within two hours post-injection, resulting in high image contrast. Such an agent may be useful to detect minute metastatic lesions. A ⁶⁴Cu-labeled dual-targeting agent that binds to PSMA and gastrin-releasing peptide receptors (GRPr) was recently reported using Glu-Glu urea as the targeting moiety for PSMA and BBN(7–14)NH₂ for GRPr with NODAGA as a chelating agents ^{93 64}Cu-Labeled PSMA-targeted agents using GPI as the targeting moiety and (CB-TE2A) as the chelator are also known.⁹⁴ Moreover, l,4,7-triazacyclononane-l,4,7-tris(2-carboxyethyl-methylenephosphinic acid (TRAP) has been used as a chelating agent/scaffold for ⁶⁴Cu and ⁶⁸Ga-labeled PSMA-targeted PET agents. A trimeric PSMA-targeted TRAP(DUPA-Pep)₃ was shown to provide higher image contrast than a monomeric analog, although renal uptake of the former was higher.⁹⁵

PSMA-targeted optical and MR imaging agents

Optical agents

In vivo optical imaging with; probes that emit in the near-infrared (NIR) region of the electromagnetic spectrum is a rapidly emerging technology. Near-infrared fluorescence imaging is highly sensitive, non-invasive, inexpensive, and has significant potential to enhance real-time, image-guided surgery, specifically for assuring negative surgical margins.^{96, 97} A wide variety of low-molecular-weight, PSMA-targeted optical imaging agents have been reported and are summarized in Table IV. Humblet et al. reported the first PSMA-targeted mono- and multivalent NIR GPI derivatives.^{98, 99} Rapid clearance of the monomeric compound from the blood limited the time required for productive engagement of PSMA-expressing tumor,⁹⁸ and in vivo results were not reported for the corresponding multivalent compounds.⁹⁹ Liu et al. have labeled PSMA inhibitors that employed the phosphoramidate and phosphate scaffold and demonstrated their PSMA-binding specificity and intracellular localization in vitro, but in vivo imaging results have not been reported.^{40, 100, 101} Banerjee and coworkers synthesized an early urea-based fluorescent compound, ReL2 (Table IV, Entry 11), but in vivo images were not obtained as the fluorophore would have an insufficient depth of penetration to be detected.⁷³ Utilizing the same Glu-Lys-urea scaffold, Chen and coworkers reported YC-27 (Table IV, Entry 12), which enabled visualization of PSMA-expressing xenografts in mice.¹⁰² They extended that work by synthesizing and testing a series of compounds using different fluorophores and linkers (Table IV, Entries 12–26). The imaging results further confirmed that a linker of appropriate length between the PSMA-binding urea and the bulky fluorophore was critical for productive PSMA targeting in vivo for this class of agents.¹⁰³ Neuman et al. have recently used YC-27 to demonstrate the utility of NIRF for assurance of negative margins in pre-clinical models of PC¹⁰⁴ (Figure 5). Kularatne,⁷⁵ Kelderhouse ¹⁰⁵ and Wang ¹⁰⁶ have synthesized a series of optical agents utilizing the Glu-Glu-urea scaffold with different linkers and demonstrated that these compounds also bind selectively to PSMA. Other preclinical agents with an optical component including hetero-bivalent compounds targeting both PSMA and integrin α_vβ₃ (Table IV, Entry 27),¹⁰⁷ as well as dual modality agents which enable sequential SPECT and optical ¹⁰⁸ and bionized nanoferrite (BNF) nanoparticles for optical and SPECT imaging.¹⁰⁹

Table IV.

Optical agents targeting PSMA

Inhibitor	Entry	Linker	Fluorophore	Reference
	1	No linker	IRDye78	98
	2	No linker	IRDye800CW	99
	3		IRDye800RS	99
	4		IRDye800CW	99
	5
	6
	7	No linker	5-FAM-X	40
	8	No linker	5-FAM	100
	9	PEG3	5-FAM	100
	10	No linker	Cy5.5	101
	11			73
	12		IRDye800CW	102
	13		IRDye800RS	103
	14		ICG derivative	103
	15		Cy7	103
	16		Cy5.5	103
	17	No linker	IRDye800CW	103
	18		IRDye800RS	103
	19		ICG derivative	103
	20		Cy7	103
	21		Cy5.5	103
	22		IRDye800CW	103
	23		IRDye800RS	103
	24		ICG derivative	103
	25		Cy7	103
	26		Cy5.5	103
	27		IRDye800CW	107
	28		FITC	75
	29		Rhodamine B	75
	30		DyLight 680	105
	31		Alexa Fluor 647	105
	32		IRDye800CW	105
	33	Amc-Ahx-Glu-Glu-Glu-Lys	IRDye800CW	106
	34		5(6)-FAM	48

Figure 5.

Near-infrared fluorescence) (NIRF) imaging with YC-27 Left panels show animals that underwent surgery with assistance of the Pearl imaging system to assure negative margins after receiving the PSMA-targeted NIRF agent, YC-27, while those on the right did not. Note re-growth of tumor as soon as seven days Post-operatively in the absence of NIR guidance (red circle, right panels) [Adapted from Neuman BP et al].¹⁰⁴

MR agents

MR molecular imaging can combine the ubiquity of this established clinical modality and its high spatial resolution with molecular profiling in vivo. However, due to the intrinsically low sensitivity of MR, nanoparticles are often employed as a platform to improve sensitivity, specifically for imaging receptors. The use of PSMA-targeted iron oxide nanoparticles as T₂-contrast agents has been investigated by several groups recently by employing either aptamer or antibody targeting.^110–115 Abdolahi and coworkers evaluated J591 -conjugated superparamagnetic iron-oxide magnetic nanoparticles (SPIONs) that displayed specific T₂ MR contrast enhancement of PSMA+ LNCaP cells but not of PSMA- DU145 cells.¹¹³ Magnetic nanoparticles (MNPs) conjugated to the PSMA-targeted antibody J591 were shown to invoke PSMA-specific MR contrast enhancement in a preclinical model of orthotopic prostate cancer.¹¹⁴ Administration of J591-MNP conjugates resulted in significant darkening of the T₂-weighted MR images within the region of the prostate at two and 24 hours post-injection compared to mice given untargeted MNPs. Recently, a PSMA-targeted polypeptide, CQKHHNYLC C1–C9 disulfide, was conjugated to the surface of a new SPION formulation for PSMA-specific MR imaging.¹¹⁵ In LNCaP tumor-hearing mice injected with polypeptide-SPIONs, T₂, signal reduction within tumors was observed from two to 12 hours post-injection. No appreciable tumor signal changes were observed in the control groups at any time point. As opposed to those nanoparticle-based T₂ contrast agents, Banerjee et al. have recently reported Gd(III)-based low-molecular-weight T₁-contrast agents weighted to provide positive contrast enhancement.¹¹⁶ Three high-affinity, low-molecular-weight gadolinium(Gd)(III)-based PSMA-targeted contrast agents containing one to three Gd-chelates per molecule (Gdl–Gd3) were synthesized employing a PSMA-targeted Glu-Lys-urea-linker construct. They evaluated the relaxometric properties of those agents in solution, in prostate cancer cells and in an in vivo experimental model to demonstrate the feasibility of T₁-weighted, PSMA-based MR contrast enhancement (Figure 6). In vivo MR imaging of Gd3 in mice bearing PSMA+ PC3 PIP and PSMA- PC3 flu tumors showed ~36% enhancement in PSMA+PC3 PIP tumor at 30 minutes post-injection and remained high with ~25% enhancement at three hours, whereas PSMA- PC3 flu tumor showed rapid decay in signal intensity after an initial ~24% enhancement at 30 minutes at 9.4 T. Other mice dosed in the same fashion using a trimeric Gd-probe without a targeting moiety,¹¹⁷ showed no significant tumor enhancement at similar time points. They concluded that PSMA is expressed in sufficient quantities on cells to enable MR-based targeting, using only modest signal enhancement built into the imaging agent.

Figure 6.

Compound Gd3 MR Imaging of human PSMA+ PC3 PIP and PSMA– PC3 flu tumor xenografts in male NOD/SCID mice. Enhancement (ΔR₁%) maps in PSMA+ PC3 PIP and PSMA– PC3 flu tumors are superimposed upon T₂-weighted images during 40–160 min after a single bolus injection of Gd3 (A). ΔR₁% maps in PSMA+ PC3 PIP and PSMA– PC3 flu tumors of a trimeric Gd contrast agent without a PSMA-targeting moiety (B). [Adapted from Banerjee SR et al].¹⁰⁴

PSMA-targeted detection of human prostate cancer

Diagnostic imaging of prostate cancer by targeting PSMA is not a new concept. The first radiotracer to target PSMA and find clinical applicability was ¹¹¹In-capromab pendetide (ProstaScint^®, Cytogen Corporation, Princeton, NJ, USA), a mouse monoclonal antibody conjugate that can be imaged using SPECT, the target epitope of which is an intracellular domain of PSMA.^118–120 Although ProstaScint^® imaging has been used extensively in the context of recurrent metastatic prostate cancer that was occult on conventional anatomic imaging, ultimately the necessity for PSMA-expressing prostate cancer cells to be lysed with exposure of the cytoplasmic target epitope,^118–120 as well as the limited spatial resolution of the SPECT acquisition and relatively high blood pool activity of the radiolabeled antibody, limited the utility of this agent.

New monoclonal antibodies and antibody derivatives targeted to PSMA have also been examined. The most extensively studied of those is the J591 antibody to the extracellular domain of PSMA. An analysis of a total of 53 patients imaged with either ¹¹¹In- or ¹⁷⁷Lu-labeled J591 found that 98% of patients with demonstrable metastatic prostate cancer had lesions that were targeted by the antibody.¹²¹ A ⁸⁹Zr-labeled conjugate of the J591 antibody has also been synthesized for use as a PET radiotracer for imaging prostate cancer. A recent prospective study in patients awaiting prostatectomy found that this agent was able to identify tumors with a Gleason score of 7 or greater.¹²² A phase I imaging trial with a ⁸⁹Zr-labeled, engineered fragment of J591 (⁸⁹Zr-huJ591) in 10 patients with metastatic prostate cancer demonstrated the ability of this construct to identify metastases, including lesions that were not noted on conventional imaging.¹²³

One of the primary shortcomings of antibody-based imaging is the long time from injection to optimal imaging time, with several days often required. That in turn limits the choice of usable radioisotopes to those with relatively long half-lives. To address that issue, multiple small-molecule ligands targeting PSMA have also been explored, as detailed in the previous sections. Among the early small-molecule PSMA ligands tested in human subjects were a pair of ¹²³I-labeled agents, ¹²³I-MIP-1072 and ¹²³I-MIP-1095, which have been found to localize to sites of metastatic prostate cancer in both bone and soft tissue.⁷¹ Agents ^99mTc-MIP-1404 and ^99mTc-MIP-1405 (Table II, Entry 10) were investigated in six healthy men and six men with radiographic evidence of metastatic prostate cancer.¹²⁴ Compared to standard MDP bone scanning, both agents identified most bone metastatic lesions and rapidly detected soft-tissue lesions including sub-centimeter lymphnodes.

Cho et al. used [¹⁸F]DCFBC, a first-generation ¹⁸F-labeled urea, in a first-in-man study of PSMA-targeted PET.¹²⁵ [¹⁸F]DCFBC enabled detection of sites of metastatic prostate cancer in bone and lymph nodes. [¹⁸F]DCFBC has subsequently been shown to be capable of imaging primary prostate cancer with radiotracer uptake correlating to Gleason Score,¹²⁶ which might allow for better selection of patients who should undergo surgery or radiation versus those who should undergo active surveillance. That agent has been found to image metastatic prostate cancer reliably in both hormone-sensitive and castration-resistant patients. Szabo et al. have also recently published a first-in-man experience with a second generation, higher-affinity PSMA-targeted radiotracer, [¹⁸F]DCFPyL, which demonstrated lower blood pool activity, higher tumor uptake, and markedly improved tumor to background ratios in comparison to [¹⁸F]DCFBC ¹²⁷ (Figure 7). [¹⁸F]DCFPyL showed maximum tumor standardized uptake values (SUV_max) as high as 100 at 2 to 2.5 hours post-injection. Another radiofluorinated agent, (2RS,4S)-2-¹⁸F-fluoro-4-phosphonomethyl-pentanedioic acid (BAY1075553), was found by Beheshti and coworkers to be able to detect primary and metastatic sites of prostate cancer, although there were reported limitations in both sensitivity and specificity.¹²⁸

Figure 7.

Maximum intensity projection images of patients with metastatic prostate cancer imaged with [¹⁸F]DCFBC (A) and [¹⁸F]DCFPyL (B). In both images, arrows demarcate sites of metastatic lymph nodes in the pelvis or retroperitoneum and arrowheads demonstrate sites of bone metastases (cervical spine in [A] and thoracic spine and left femoral head in [B]]. Note the markedly reduced blood pool activity and overall higher tumor-to-background ratio with [¹⁸F]DCFPyL.

Outside of those ¹⁸F-labeled agents, a great deal of clinical experience has been gained in Europe utilizing a ⁶⁸Ga-labeled, urea-based compound, referred to as ⁶⁸Ga-PSMA, i.e., [⁶⁸Ga]DKFZ-PSMA-11 (Table III, Entry 10).⁸³, ^{129 68}Ga has the distinct advantage of being generator-produced, allowing it to be used at sites without ready access to a medical cyclotron. However, relative to ¹⁸F, ⁶⁸Ga has a shorter physical half-life (68 minutes versus 110 minutes) that limits its ability to be produced at a central location and shipped to other sites and ⁶⁸Ga also produces higher energy positrons that, in theory, degrade its spatial resolution.¹³⁰ Nevertheless, a number of studies and case reports have been published that demonstrate that the ⁶⁸Ga ligands are able to detect prostate and other cancers reliably in a variety of contexts. A study of 37 patients with biochemically recurrent prostate cancer demonstrated a higher lesion detection rate with ⁶⁸Ga-PSMA PET/CT in comparison to ¹⁸F-fluoromethylcholine PET/CT.¹³¹ A large retrospective analysis of 319 patients with biochemical recurrence found that ⁶⁸Ga-PSMA was able to detect sites of disease in 82.8%, with reported sensitivity, specificity, negative predictive value, and positive predictive value on a lesion-by-lesion basis of 76.6%, 100%, 91.4%, and 100%, respectively.¹³² A separate retrospective analysis of 248 patients with biochemical recurrence demonstrated similar findings, with 89.5% of patients being found to have radiotracer-avid sites of disease.¹³³ In addition to those larger studies, case reports have been published demonstrating the ability of ⁶⁸Ga-PSMA to detect a brain metastasis due to prostate cancer,¹³⁴ to be used with MR imaging for the detection of primary disease,^{135, 136} and to evaluate the response of bone metastases to ²²³Ra therapy.¹³⁷ Although both ¹⁸F- and ⁶⁸Ga-labeled small-molecule radiotracers for PSMA have demonstrated excellent promise, a recent manuscript by Dietlein et al. suggested that small lesions in patients with biochemical recurrence may be evaluated with higher fidelity by [¹⁸F] DCFPyL than by ⁶⁸Ga-PSMA.¹³⁸

PSMA-targeted therapy and theranostic agents in human prostate cancer

Targeting of PSMA for prostate cancer therapy has been investigated using antibodies, minibodies, antibody-drug conjugates (ADCs), and small molecules. We will briefly discuss the former agents and then focus on small-molecule agents.

Antibody-based therapies

Antibody therapy targeting PSMA was initiated using radiolabeled capromab pendetide (CYT-356) in two phase 1 dose-escalation studies, one in patients with metastatic castration-resistant prostate cancer and the other in patients with biochemical recurrence.^{139, 140} In patients treated with ⁹⁰Y-labeled CYT-356, however, efficacy was lacking as antibody was unable to bind to viable tumor cells, and patients experienced dose-limiting myelosuppression. Subsequent clinical trials have investigated the deimmunized IgG monoclonal antibody, J591, previously radiolabeled with ¹¹¹In, ⁹⁰Y, and ¹⁷⁷Lu through a DOTA chelator. Patients that underwent trials with ⁹⁰Y- or ¹⁷⁷Lu-labeled J591 experienced reversible myelosuppression at higher doses.^{121, 141} Radiographic tumor response and significant prostate-specific antigen (PSA) decline were observed only in patients receiving ¹⁷⁷Lu-labeled J591. Based on those results, a phase II trial in metastatic castration-resistant prostate cancer was initiated at two centers, with an initial cohort receiving 65 mCi/m² and a second cohort receiving 70 mCi/m². Imaging studies revealed excellent targeting (94% of patients) and PSA declines were significant and greater in the higher dose cohort (71% vs. 46%), with thrombocytopenia as the most common severe hematologic toxicity and no significant non-hematologic toxicities.¹⁴² A subsequent phase I trial of fractionated-dose ¹⁷⁷Lu-labeled J591 demonstrated improved tolerability.¹⁴³ Phase I trials using fractionated ¹⁷⁷Lu-J591 for radioimmunotherapy plus docetaxel as a radiosensitizer are ongoing [NCT00916123].

A significant drawback of antibodies for therapy is that full antibodies do not homogenously penetrate tumors, resulting in diminished therapeutic potential. Therefore, Watanabe et al. developed an approach using antibody fragments such as small bivalent antibody fragments, or minibodies.¹⁴⁴ In a preclinical study they found that NIR photoimmunotherapy could be effected at an earlier time point after administration of the PSMA-targeted minibody relative to the intact antibody. Such studies further underscore the importance of pharmacokinetic optimization of PSMA-targeted agents to enable practical clinical implementation.

Another antibody-mediated approach has been to engineer J591 to interact with human immune effector cells to trigger antibody-dependent cell-mediated cytotoxicity (ADCC). In a dose-escalation trial in patients with progressive castration-resistant prostate cancer, dose correlated with degree of ADCC induction, with one patient who received 100 mg of J591 experiencing a reduction in PSA of greater than 50%.¹⁴⁵ In a phase II trial combining IL-2 (known to promote immune responses) and J591, there was a trend toward less progression in patients demonstrating natural killer cell expansion.¹⁴⁶

Antibody-drug conjugate therapies

Significant interest has surrounded enhancing the efficacy and/or reducing the toxicities of cytotoxic chemotherapies by using PSMA-targeted ADCs or nanoparticles. A pre-clinical study that employed a polymeric nanoparticle targeting PSMA containing the chemotherapeutic agent docetaxel (DTXL-TNP) showed enhanced tumor accumulation at 12 hours and prolonged tumor growth delay compared to solvent-based DTXL.¹⁴⁷ Preliminary reports of an ongoing phase 1 study in patients with advanced or metastatic solid cancers showed tumor shrinkage with DTXL-TNP at doses lower than solvent-based DTXL.¹⁴⁷ MLN2704, an ADC created by conjugating J591 to the anti-microtubule drug maytansinoid-1, was evaluated in a phase I study after demonstrating preclinical activity.^{148, 149} Although a subsequent phase I/II study demonstrated limited efficacy with significant neurotoxicity using MLN2704, the trial established the potential of PSMA-targeted ADCs.¹⁵⁰ As a result, an additional PSMA-based ADC directing enzyme-activated cytotoxin release, and another delivering the toxin monomethylauristatin E (MMAE), are undergoing phase I and II investigations, respectively.^{151, 152}

Small molecule therapies

Recent studies have demonstrated the potential for clinical translation of PSMA-targeted small-molecule radiopharmaceutical therapies, as well as theranostic approaches combining targeted imaging and targeted therapy. Structures of recently reported ⁶⁸Ga/¹⁷⁷Lu-based PSMA-targeting theranostic radiometals are shown in Table III, entries 2–15. Weineisen and coworkers recently evaluated three compounds,^{153, 154} in which, the DOTA-monamide chelating agent (Table III, Entry 2) was replaced by 1,4,7,10-tetraazacyclodocecane-1-(glutaric acid)-4,7,10-triacetic acid (DOTAGA), while maintaining the same Glu-Lys urea PSMA-targeting construct. The authors reported that due to metabolic instability of the L-amino acid residues on the linker, D-amino acids were implemented to provide metabolically stable agents. [⁶⁸Ga]DOTAGA-ffk(Sub-KuE) (Table III, Entry 11) displayed favorable pharmacokinetics and high tumor accumulation in PSMA+ LNCaP tumor-bearing mice.¹⁵³ Further modification of the linker, specifically, replacing one D-Phe residue with lipophilic 4-iodo-D-tyrosine, resulted in [⁶⁸Ga]DOTAGA-(Iy)fk(Sub-KuE), i.e., PSMA I&T (Table III, Entry 12), which was recently evaluated both in a murine model and in initial proof-of-concept clinical studies.¹⁵³ Baum et al. have used a ¹⁷⁷Lu-labeled version of PSMA I&T in 60 patients with significant decreases in PSA observed and without demonstrable side effects. One patient experienced a decrease in PSA from 36 ng/mL to <1 ng/mL after three cycles of therapy (Figure 8) (Richard Baum, personal communication). The compounds of Table III, Entries 13 and 14, prepared by using CHX-A″-DTPA and DO3A, respectively, employed the Glu-Glu urea as the targeting moiety.^{155, 156}

Figure 8.

⁶⁸Ga-PSMA PET/CT following therapy with [¹⁷⁷Lu]PSMA I&T. Patient presented with PSA=35.9 ng/mL and a PET scan showing multiple, intense mediastinal, abdominal and pelvic lymph node metastases, as well as lesions in the thoracic spine and local recurrence in the prostate fossa (A). Ten months later, after three cycles of peptide radioligand therapy (PRLT) PSA decreased to 1.0 ng/mL with significant regression of previously noted disease (B). Three foci of radioactivity superimposed upon the thorax are due to retention of radioactivity within the Port-A-Cath^® device in B. Images courtesy Dr. Richard Baum, Zentralklinik Bad Berka, DE.

The radiohalogenated agent [¹³¹I]MIP-1095 (Table I, Entry 8) has been investigated in 28 patients with metastatic castration-resistant prostate cancer. After receiving a single cycle of [¹³¹I]MIP-1095 PSA values decreased by >50% in 61% of the men treated, with transient dry mouth as the most notable side-effect (in 25% of patients). With that treatment, involved lymph nodes and bone metastases received an estimated dose of more than 300 Gy of targeted radiation therapy.⁷² Furthermore, the Auger electron emitter ¹²⁵I has been conjugated to urea-based DCIBzL (Table I, Entry 5) and has recently been shown to yield significant antitumor efficacy in mice bearing PSMA+PC3 PIP xenografts, emphasizing the importance of PSMA internalization and perinuclear localization, as Auger electrons are known to have a very short range of <10 μm.¹⁵⁷

Preclinical theranostics

Many of the translational theranostic agents targeting PSMA that employ radiometals use a strategy similar to that depicted in Figure 4 where the linker has been modified with various hydrophobic aromatic species in order to enhance affinity as first reported by Kularatne et al.⁷⁵ With respect to possible theranostics, Banerjee et al. have explored the positron emitting radiometal ⁸⁶Y (T_1/2=14.7 h) as an imaging radionuclide for PSMA, as agents incorporating it can be used for dosimetry estimates of the corresponding pure β-particle emitter, ⁹⁰Y, and possibly for modeling compounds that incorporate ¹⁷⁷Lu, which also uses the DOTA chelator. Structures are presented in Table III, Entries 4 and 7. Compound [⁸⁶Y]6 (Table IV, Entry 7) was generated using the p-Bn-SCN-DOTA chelator and a diaminobutane linker, and provided the desired lower kidney uptake and higher tumor retention than other agents in this series,¹⁵⁸ making this agent particularly attractive for radiotherapy.

More complex PSMA-targeted theranostic agents, namely, those that do not solely leverage radionuclide therapy, have so far not been translated to the clinic. Chen et al. published a series of proof-of-principle experiments showing the versatility of a theranostic nanoplex.¹⁵⁹ The nanoplex contained three covalently linked components: 1) a pro-drug-activating enzyme bacterial cytosine deaminase (which converts a nontoxic pro-drug to the cytotoxic 5-fluorouracil); 2) an imaging reporter carrier poly-L-lysine labeled with the NIRF probe Cy 5.5; and 3) a vector for siRNA delivery to down-regulate choline kinase (which may enhance the effect of 5-FU). The nanoplex was radiolabeled with [¹¹¹In] DOTA for SPECT imaging and targeted to PSMA through a low-molecular-weight, urea-based moiety. That strategy can be extended to other targets and types of cancer, increasing the precision of drug delivery to enhance the impact and safety of therapy, the chief goals of nanomedicine.

PSMA as a target for imaging and therapy in other cancers

PSMA expression in renal tubules, duodenal and colonic mucosa, and brain has been well defined.¹⁶⁰ PSMA expression on neovascular endothelial cells in a variety of solid tumors is likewise known, as indicated above. Literature on PSMA over-expression in non-prostate prostate tumor epithelial cells is beginning to emerge. In this section, we will discuss PSMA over-expression in epithelial and endothelial cells of non-prostate tumors and provide an update on its use for imaging and therapy in this context.

In one of the earliest observations, Silver et al. reported intense PSMA staining in the endothelial cells of capillary vessels in peritumoral and endotumoral areas of renal and colon carcinomas.¹² Further confirmatory studies to detect PSMA expression in neovasculature were later carried out by Chang et al., in benign tissues, prostate and 15 types of non-prostate tumors using five different PSMA monoclonal antibodies. Those studies showed strong PSMA immunoreactivity in the neovasculature of a variety of tumors including renal cell carcinoma (RCC), glioblastoma (GBM), melanoma and non-small cell lung cancer.³² No staining was observed in the epithelial cells of non-prostate tumors. Although tumor epithelial cells were universally positive for PSMA, interestingly only a fraction of the corresponding tumors showed PSMA immunoreactivity in the neovasculature.³² Those immunohistochemical observations were confirmed using RT-PCR and in situ hybridization studies in different tumor types by the same group.³¹

PSMA expression is well characterized in the neovasculature of renal neoplasms. Nearly 76% of clear cell RCC and 31% of chromophobe RCC were shown to be positive for PSMA.¹⁶¹ PSMA expression in RCC may become important for this cancer with few diagnostic imaging options. Also, PSMA expression was detected in 66% of gastric and 85% of colorectal carcinomas tested.³⁴ Similarly, PSMA expression in neovasculature from GBM was demonstrated to a variable extent in 100% of specimens tested.¹⁶² Results from other groups confirmed those observations,¹⁶³ suggesting a new role for PSMA-targeted imaging and therapeutics in management of GBM. PSMA expression was shown to be associated with poor survival outcome for several non-prostate tumor types. Detection of PSMA within neovasculature in nearly 50% of oral squamous cell carcinoma specimens tested was associated with reduced survival.¹⁶⁴ Similarly, 46% of osteosarcoma specimens showed PSMA immunoreactivity in the neovasculature that correlated with poor survival.¹⁶⁵ In a recent study by Wernicke and coworkers, 74% of the primary breast cancers tested showed PSMA immunoreactivity in the neovasculature.¹⁶⁶ High PSMA expression in the neovasculature on breast cancers was shown to result in reduced overall survival.

In addition to primary tumors, PSMA expression was also detected in the neovasculature of metastatic lesions. Metastatic tissue from RCC showed consistent PSMA expression in the neovasculature.³³ Also, nearly 80% of liver and colon carcinoma metastases tested were positive for PSMA in the neovasculature.¹⁶⁷ Similarly, 100% of brain metastases due to breast cancer showed PSMA immunoreactivity in the neovasculature.¹⁶⁶ Availability of PSMA-targeted antibodies and their detailed characterization has improved the detection of PSMA in the epithelial cells of non-prostate tumors. PSMA expression in lung cancer cells was recently reported. Nearly 54% of lung tumor cells and 85% of lung tumor endothelial cells demonstrated PSMA immunoreactivity.¹⁶⁸ Although PSMA expression in non-prostate tumor epithelial cells and endothelium is gaining traction, and case reports are appearing, its diagnostic and therapeutic value in non-prostate cancer has yet to be established.

In a recent study, Pandit-Taskar et al. investigated the use of ¹¹¹In-labeled J591 for imaging a variety of solid tumors, demonstrating that 74% of skeletal lesions, 53% of lymph node lesions, and 64% of other visceral lesions were identified by the antibody, with uptake seen in lesions from all investigated tumor types.¹⁶⁹ Also, Demirci et al. imaged a patient with known metastatic ccRCC using ⁶⁸Ga-PSMA PET/CT.¹⁷⁰ The authors observed multiple sites of metastatic disease that were occult with FDG PET/CT. Rowe et al. have similarly observed metastases due to ccRCC using [¹⁸F]DCFPyL (unpublished results). Comparison with FDG PET/CT indicated higher standardized uptake values for the PSMA-targeted agent in this tumor that is typically difficult to detect with FDG (Figure 9). Those results are important due to the often small and unusual sites of metastasis seen with ccRCC that point to a potential increase in diagnostic yield with PSMA-based imaging, as well as the possibility of evaluating response to therapy given the anti-angiogenic action of many of the chemotherapeutic options for metastatic ccRCC (e.g. various tyrosine kinase inhibitors).

Figure 9.

Axial FDG PET/CT (A) and axial [¹⁸F]DCFPyL PET/CT (B) images of a patient with metastatic clear cell renal cell carcinoma involving the posterior aspect of the right iliac bone. Both images are set to the same quantitative scale, highlighting the increased tumor uptake achieved with [¹⁸F]DCFPyL in this case.

Conclusions

Dedicated work over the last decade in medicinal chemistry has provided a number of small-molecule SPECT and PET ligands for in vivo imaging of PSMA, as well as theranostic and therapeutic agents. A number of those agents have now been translated and clinically evaluated in patients with prostate cancer, and specific and appropriate indications for their application are emerging. Given the higher sensitivity and specificity of PSMA-based PET in the detection of metastatic prostate cancer relative to contrast enhanced CT and bone scan, it seems that these agents will be well suited for the evaluation of patients with early biochemical recurrence. Indeed, in such patients the ability to identify recurrent metastatic disease within the pelvis as opposed to disease at more distant sites would allow for selection of appropriate therapy, such as salvage pelvic radiotherapy versus systemic androgen deprivation therapy. The use of PSMA imaging agents for upfront staging of patients with high-risk disease may allow for detection of unsuspected sites of lymph node or bone metastases, again guiding selection of therapy. In those patients with low-risk, organ-confined disease, early data suggest that uptake of PSMA-targeted agents may correlate with the aggressiveness of the primary tumor, which may aid in selection of patients who should undergo surgery or radiotherapy versus active surveillance.

Emerging data suggest that radiolabeled compounds targeting PSMA may prove to be effective theranostic and therapeutic options for patients with metastatic prostate cancer. Promising early clinical studies with ⁶⁸Ga/¹⁷⁷Lu-based PSMA-targeting radiometals have demonstrated tumor responses with surprisingly little toxicity. Such compounds may find an important role in treatment, particularly in patients whose tumors have become castration-resistant but retain PSMA expression. Furthermore, the expression of PSMA in the neovasculature of numerous non-prostate solid tumors coupled with recent reports of successful PSMA-based PET/CT imaging of such cancers support the concept that PSMA-targeted imaging and therapy can be extended beyond management of prostate cancer.