18F-Labeled, PSMA-Targeted Radiotracers: Leveraging the Advantages of Radiofluorination for Prostate Cancer Molecular Imaging

Prostate-specific membrane antigen (PSMA)-targeted PET imaging for prostate cancer with 68Ga-labeled compounds has rapidly become adopted as part of routine clinical care in many parts of the world. However, recent years have witnessed the start of a shift from 68Ga- to 18F-labeled PSMA-targeted compounds. The latter imaging agents have several key advantages, which may lay the groundwork for an even more widespread adoption into the clinic. First, facilitated delivery from distant suppliers expands the availability of PET radiopharmaceuticals in smaller hospitals operating a PET center but lacking the patient volume to justify an onsite 68Ge/68Ga generator. Thus, such an approach meets the increasing demand for PSMA-targeted PET imaging in areas with lower population density and may even lead to cost-savings compared to in-house production. Moreover, 18F-labeled radiotracers have a higher positron yield and lower positron energy, which in turn decreases image noise, improves contrast resolution, and maximizes the likelihood of detecting subtle lesions. In addition, the longer half-life of 110 min allows for improved delayed imaging protocols and flexibility in study design, which may further increase diagnostic accuracy. Moreover, such compounds can be distributed to sites which are not allowed to produce radiotracers on-site due to regulatory issues or to centers without access to a cyclotron. In light of these advantageous characteristics, 18F-labeled PSMA-targeted PET radiotracers may play an important role in both optimizing this transformative imaging modality and making it widely available. We have aimed to provide a concise overview of emerging 18F-labeled PSMA-targeted radiotracers undergoing active clinical development. Given the wide array of available radiotracers, comparative studies are needed to firmly establish the role of the available 18F-labeled compounds in the field of molecular PCa imaging, preferably in different clinical scenarios.


Introduction
As one of the most common malignancies in men in the United States and Europe, the incidence of metastatic prostate cancer (PCa) continues to rise [1][2][3]. The ability to ascertain the presence and extent of metastatic disease has been significantly increased through the introduction of imaging agents targeting prostate-specific membrane antigen (PSMA) which provide superior diagnostic performance compared to alternative techniques, with a calculated positive predictive value of >95% [4][5][6]. These PSMA-targeted PET radioligands have not only been found to have diagnostic accuracy for visualizing sites of PCa, but have also paved the way for a theranostic approach utilizing alpha/beta-labelled agents, in a manner similar to peptide receptor radionuclide therapy for neuroendocrine tumors [7][8][9][10]. In this regard, the theranostic principle is based on the concept of a predictive biomarker (e.g. sufficient uptake on a PSMA-targeted PET scan in putative sites of disease), followed by an indivualized treatment with a therapeutic agent, e.g. by targeted ßtreatment using 177 Lu-labeled PSMA inhibiting compounds [11,12].
In general, a PET-based radionuclide for imaging PCa should inherit several key properties to generate good imaging quality and to satisfy the needs of the treating urologist: the compound should (I.) have a high affinity towards a target on PCa cells providing very high tumor-to-background ratios, (II.) identify even very small volume sites of disease, (III.) have a biodistribution favorable to the detection of typical sites of disease such as the prostate bed, pelvic lymph nodes, and bone, (IV.) work reliably among a large variety of different clinical contexts including initial staging, restaging at the time of biochemical recurrence, radiation or surgery planning, (V.) have a high radiochemical yield enabling for a high throughput of patients, (VI.) have been extensively validated in a (pre-)clinical setting, (VII.) outperform current state-of-the-art imaging approaches, and (VIII.) allow for endoradiotherapy by labelling structurally related agents with 177 Lu, 90 Y, or alpha-emitting radionuclides.
To date, 68 Ga-labeled imaging agents are by far the most commonly used PSMA-targeted PET radiotracers in clinical practice [13]. However, recent years have witnessed the beginning of a shift from 68 Ga-to 18 F-labeled compounds [14]. The latter radionuclide offers several advantages compared to 68 Ga including: a.) flexibility in study design with potential for delayed imaging protocols due to a longer half-life, b.) lower positron energy with resultant short positron range in tissue and improved contrast and noise, and c.) potential for distribution by commercial vendors through existing shipping networks [15,16]. Notably, such dispatch systems from central cyclotron facilities to smaller hospitals have led to remarkable cost-savings for [ 18 F]FDG [17]. Thus, if a more widespread adoption of PSMA-targeted PET imaging is to be pursued in the near future, 18 F-labeled radiotracers may play a pivotal role in meeting this increasing demand in patient care. In addition to that, recent studies have already suggested that the theoretical improved imaging possible with 18 F-compared to 68 Ga-labeled PSMA agents may already have an impact on patient care, such as in subjects with biochemically-relapsed PCa [18][19][20]. Nonetheless, 18 F labeling chemistry has also several drawbacks, e.g. time-consuming labeling procedures (e.g. indirect labeling methods instead of direct fluorination methods using a single precursor) [21], challenging radiosynthesis (e.g. varying chemoselectivity for the incorporation of 18 F into compounds such as peptides) [22], installation and maintenance of a costly cyclotron, and limited experience with theranostic approaches based on 18 F-labeled compounds as the diagnostic agents [23].
However, given the increasing use of 18 F radiochemistry for molecular imaging in PCa, we aimed to provide an overview of recently introduced 18 F-labeled radiotracers for PSMA-targeted PET imaging.
Older, first-generation, 18 F-labeled radiotracers for PSMA including [ 18 F]DCFBC and the phosphonomethyl compound BAY 1075553 will not be the focus of this review of more-recentlydeveloped agents, but we mention them here to acknowledge the important role of such compounds in providing early evidence of the feasibility of imaging PSMA with 18 F-labeled agents [24][25][26][27][28][29][30]. This review will focus on newer radiotracers including clinical development of a number of agents, such as the clinically established radiotracers [ 18 [14,[31][32][33][34][35]. Figure 1 illustrates the chemical structures of selected 18 F-labeled compounds [32].

F vs. 68 Ga radiochemistry in molecular imaging of PCa
Compared to 18 F, 68 Ga has multiple drawbacks: 1.) The lower positron yield ( 68 Ga, 89.14% vs. 18 F, 96.86%) and higher positron energy ( 68 Ga, 1,899 keV vs. 18 F, 633 keV) of 68 Ga may impact image quality and degrade diagnostic yield [36]. Among other physical properties, those factors contribute to the partial volume effect, which in turn has a major impact on semi-quantification, e.g. by intra-or inter-lesion comparisons using standardized uptake values (SUV) [37]. For instance, the image spatial resolution is inferior with 68 Ga compared to 18 F (2.4 vs 1.4 mm in all directions) [38]. Given the presence of multiple single-photon emissions, imaging noise is also increased with 68 Ga [37].
2.) In light of the large variety of commercially-available 68 Ga generators, varying properties of these generators (e.g. hydrochloric acid concentration, eluate processing procedures, or cationic/anionic exchange cartridges) have to be taken into account [39]. The ability to mass produce 18 F-labeled PSMA inhibitors will provide a supply of PET agent that would saturate any demand for such agents. Similar supply cannot be met with a 68 Ga generator-dependent PET agent [36].
3.) Delivery of 18 F-labeled radiopharmaceuticals from distant suppliers expands the availability of PET radiopharmaceuticals in smaller hospitals operating a PET center remote from a cyclotron facility [17]. This is particularly important given the growing demand for PSMA-targeted PET scans around the globe. However, for 68 Ga generators, a relatively high throughput of patients is needed to reduce the cost per injected dose [40]. Thus, commercial sources for 18 F-labeled compounds can offer a cost-effective alternative to in-house cyclotrons, as has been proven for [ 18 F]FDG [41]. In the U.S., Medicare spending is anticipated to double between 2010 and 2030 to approximately $1.2 trillion and thus, escalating health care expenses increase the demand for such effective cost containment strategies [42]. Purchasing [ 18 F]FDG from a production supplier located 1200 miles away, Durcharme et al. have not only reported on the feasibility of incorporating such a concept in the clinic, but also summarized practical aspects when receiving 18 F-labeled agents from a distant supplier [17]. In addition, compared to 18 F, the shorter half-life of 68 Ga will limit the shipping range of 68 Ga-labeled PSMA-targeted radiotracers to remote PET facilities. 4.) The longer half-life of 18 F allows for delayed imaging protocols. Using [ 68 Ga]PSMA-HBED-CC, Hohberg and co-workers reported on a higher lesion detection rate for a later imaging time-point as compared to early imaging (1h vs 3h post-injection), in particular for patients with low PSA levels [43]. Schmuck et al. have also shown that in patients with biochemical recurrence or PSA persistence after primary therapy for PCa, tumor-to-background ratios have increased at later imaging time-points using [ 68 Ga]PSMA I&T [44]. 18 F has a longer half-life and thus, one may hypothesize whether even more delayed imaging protocols with 18 F-labeled PSMA radiotracers may also further increase overall detection rate. This also applies to increased acquisition times, which improves overall imaging quality by reducing image noise in PCa molecular imaging [45]. 5.) The success of 18 F-labeled, PSMA-targeted PET agents is most likely owed to the widespread adoption of the most commonly used radiotracer in oncology, namely [ 18 F]FDG [46]. Thus, for 18 F-labeled imaging agents for PCa, existing cylotrons, infrastructure, and protocols for production, handling, and transport of 18 F can be effectively utilized.  Lower positron yield and longer positron range lead to an increased partial volume effect, which in turn hampers diagnostic accuracy and semiquantitative approaches [37] Limited time-span between injection of the radiopharmaceutical and start of the scan [37] Increased imaging noise compared to 18 F [37] 68 Ga-based synthesis allows for the production of sufficient radioactivity for 5 to 6 scans per day (depending on the time point of the generator life span) [36] Large variety of commercially available 68 Ga generators with varying properties, e.g. different hydrochloric acid concentration or eluate processing procedures [39] Deterioriating performance of 68 Ge / 68 Ga generators Most commonly used radiotracer in clinical practice to date for prostate cancer Easily obtainable via commercially available 68 Ga generators Extensively tested in large clinical trials, e.g. in a prospective single-arm trial enrolling 635 men [47] Extensively used in a theranostic setting for subsequent 177 Lu-based therapies [12] 18 F Installation and maintenance of a costly cyclotron is required [23] Time-consuming and challenging radiosynthesis [23] No larger prospective trials to date (comparable to 68 Ga-PSMA PET [47]) Experience with theranostic approaches based on 18 F-PSMA PET are limited Theoretically, it allows for the injection of less radioactivity, which minimizes radiation exposure to both patients and personnel [15] Increases the stability of a radiopharmaceutical against metabolism at sensitive positions and allows for higher flexibility in the study design [15] Allows for delayed imaging protocols, which may increase lesion detection rate (in a manner similar to late imaging protocols using 68 Ga-labeled PSMA PET radiotracers) [15,43,44,93] Cyclotron-based synthesis of 18 F allows for the production of larger batches on the basis of a single synthesis [36] Delivery to smaller hospitals operating a PET center remote from a cyclotron facility: -Expands the armamentarium of available PET radiopharmaceuticals in smaller hospitals -Cost-effective alternative (compared to in-house productions), as shown with [ 18 F]FDG [41] Nonetheless, several drawbacks of 18 F-labeled compounds have to be considered, in particular when compared to current 68 Ga-PSMA PET compounds: 1.) The time-consuming and challenging 18 F radiochemistry is an important potential limitation [23], although centralized production of 18 F-labeled compounds, as opposed to onsite production, would substantially mitigate this drawback.
2.) Given the increasing availability of 68 Ge /68 Ga generators with a half-life of 271 days, an on-site cyclotron is not necessary for having access to 68 Ga for labeling radiopharmaceutical precursors [37].
3.) 68 Ga-labeled PSMA specific compounds are better evaluated based upon their long-standing use in daily patient care. For instance, Fendler et al. recently evaluated [ 68 Ga]PSMA-11 in a prospective single-arm trial enrolling 635 men [47].

Part I: Clinically established 18 F-labeled radiotracers [ 18 F]DCFPyL
First, Szabo and co-workers reported on nine patients with known metastatic PCa. Apart from an acceptable safety profile, increased radiotracer accumulation was appreciated in presumed primary and metastatic sites of disease, along with an effective radiation dose comparable to [ 18 F]FDG (0.0319 mGy/MBq after injection of 370 MBq). In descending frequency, the highest radiation dose was recorded in the kidneys (most likely caused by specific binding), followed by the urinary bladder wall, the submandibular glands and the liver. Compared to other 18 F-labeled PSMA agents, the latter finding of less hepatic uptake may play a role for an increased detection rate of liver lesions attributable to PCa [14] and may aid in the identification of high retroperitoneal lymph nodes.
Not surprisingly, [ 18 F]DCFPyL outperformed other molecular imaging agents ( 99m Tc-methylene diphosphonate bone scan, Na 18 F PET/CT) for lesion detection in patients with PCa. In a 45-year old man with suspected oligometastatic PCa, 87 definitive sites of non-physiologic [ 18 F]DCFPyL accumulation were detected, while planar bone scan detected only 12 supicious sites and Na 18 F PET/CT 39 lesions [50]. The same research group also reported on the superiority of [ 18 F]DCFPyL compared to conventional imaging on a larger basis. In a secondary analysis of the first nine patients imaged with [ 18 F]DCFPyL, 138 definitive sites of [ 18 F]DCFPyL uptake were recorded, while only 1 was classified as equivocal. This was in contradistinction to CT-and-bone-scan based disease patterns (30 definitive, 15 equivocal). Taking into account intra-patient clustering effects, lesions which have been rated as negative/equivocal on conventional imaging would be positive on [ 18 F]DCFPyL in the vast majority of the cases (0.72, based on a generalized estimating equation regression analysis) [5]. In addition, [ 18 F]DCFPyL has also been tested in a head-to-head comparison with its 68 Ga-labeled counterpart, namely [ 68 Ga]PSMA-11. In 14 patients with biochemically recurrent disease, the 18 F-labeled radiotracer identified more sites of disease in three patients and those findings were also further corroborated semiquantitatively (higher maximum SUV (SUV max ) and tumor-to background ratios). The results of this study suggest [ 18 F]DCFPyL is an attractive alternative to 68 Ga-labeled compounds ( Figure 2) [18].
Given In an organ-to-background analysis, [ 18 F]DCFPyL had significant higher radiotracer accumulation in the urinary system and lacrimal glands, which may hamper diagnostic accuracy in small lymph nodes in the (lower) pelvis. This was in contradistinction to [ 18 F]PSMA-1007, which demonstrated increased uptake levels in the muscle, submandibular and sublingual glands, spleen, pancreas, gallbladder, and liver. The latter organ uptake may decrease the level of certainty of a reader that an equivocal finding in the liver is a site of disease, in particular without a cross-sectional imaging correlate [51].
Wondergem et al. compared a 60 vs. 120 min time-to-scan interval using [ 18 F]DCFPyL and showed that >38% of all investigated patients had more suspicious lesions when the later imaging protocol was applied. Notably, this led to a change in TNM staging in >9% [52]. If there is widespread adoption of 18 F-labeled, PSMA-targeted radiotracers for PCa imaging or even an incorporation of those agents in clinical guidelines, such observations are of utmost importance for establishing adequate imaging protocols.
Many pitfalls for both 68 Ga-and 18 F-labeled PSMA-targeting radiotracers have been reported since their routine introduction in the clinic [54]. These include, but are not limited to: benign sites of radiotracer accumulation (e.g. ganglia which can be misinterpreted as lymph node metastases or healing fractures as bone lesions), benign vascular tumors (hepatic hemangioma), pulmonary involvement (sarcoidosis or tuberculosis) or even non-prostatic malignant tumors such as renal cell carcinoma [55]. To address these manifold pitfalls, framework systems for PSMA-targeted PET imaging have been recently introduced, and such reporting systems may meet the need to harmonize interpretation of PSMA PET-based findings [56,57]. Readers with different levels of experience each read 50 [ 18 F]DCFPyL PET/CTs and utilized the PSMA reporting and data system (PSMA-RADS) framework to classify up to five target lesions. Notably, using such a structured reporting system, the majority of the PET studies were assigned to the highest PSMA-RADS score (PSMA-RADS-4 or -5), which indicates that the readers were convinced that PCa was either highly likely or certainly present. Thus, this study further suggests the high sensitivity and specificity of [ 18 F]DCFPyL for assessing sites of disease solely on a visual level. Moreover, the high concordance rate, even among readers with different levels of experience, suggests that this radiotracer is nearing readiness to be implemented in the collection of data for larger prospective trials [58].
Beyond its potentially pivotal role in PCa imaging, an extensive body of literature has also reported on the utility of this radiotracer for non-PCa malignancies and malignant-like conditions, e.g. for renal cell carcinoma, oncocytoma, and high-grade gliomas [59][60][61][62][63]. These findings render this radiopharmaceutical an attractive compound for other tumor entities, not only for diagnostic purposes, but also for selecting patients for radioligand treatment (RLT). However, prior to an implementation as a potential therapeutic radiotracer, reproducible methods for a reliable (peri)therapeutic dosimetry are indispensable, which in turn allows for maximizing therapeutic efficacy but minimizing potential harm [64] [14,31]. Beyond a report on visual assessment, the authors also performed a histopathological comparison of patients who underwent prostatectomy and the diagnostic findings were validated by examinations of harvested tumor tissue [31]. Based on these encouraging findings, Rahbar et al. evaluated 100 patients with biochemical relapse in a retrospective setting and distinctive detection frequencies among patients with different PSA levels were noted, with rates of positive scans as follows: 86%, 89%, 100% and 100% among patients with PSA levels of ≤0.5, 0.51-1.0, 1.1-2.0 and > 2.0 ng/ml, respectively [68]. Giesel et al. provided further insights on the diagnostic efficacy of [ 18 F]PSMA-1007 in >250 patients suffering from biochemical recurrence after radical prostatectomy: Further improvements in the overall detection rate were recorded, in particular in low and ultra-low PSA levels of 0.5-<1 and 0.2-<0.5 ng/mL with detection rates of 74.5% and 61.5%, respectively [69]. Rahbar and coworkers also reported on the superior diagnostic performance of [ 18 F]PSMA-1007 in comparison with [ 68 Ga]PSMA-11, in particular for segregating local recurrence from physiological radiotracer excretion [70]. Similar to [ 18 F]DCFPyL, [ 18 F]PSMA-1007 has an increased lesion detection rate at 120 min p.i. compared to an early imaging protocol 60 min p.i., along with a significant increase in SUV max [71] ( Figure 3) For the latter agent, a higher number of lesions with quantitatively increased radiotracer accumulation attributed to benign lesions were found, which further emphasizes the need of increasing reader's confidence to identify such pitfalls [73]. Thus, incorporation of structured reporting systems may be of relevance for an accurate scan interpretation, in particular when reading an 18 F-labeled PSMA PET scan [57].  68 Ga-labeled PSMA-targeting radiotracers.
Preclinical Investigations. While both [ 18 F]DCFPyL and [ 18 F]PSMA-1007 are based around an urea moiety that binds in the PSMA active side, the novel 18 F-labeled compound CTT1057 is based on a phosphoramidate core (Figure 1). Incorporating such a phosphoramidate scaffold, the binding to PSMA is irreversible (based on IC50 determinations), which may result in improved target-to-background ratios. A structurally related compound was investigated in [PSMA+] xenografts and demonstrated good imaging qualities, along with high uptake in sites of disease and rapid non-target clearance [74]. Notably, [ 177 Lu]CTT1403 is based on the same binding scaffold like its PET counterpart CTT1057 and the therapeutic efficacy of the 177 Lu-labeled compound has already been tested in a preclinical setting [75].    [14,31,76]. In 5 patients that were scheduled for imaging with [ 18 F]CTT1057 followed by prostatectomy, 4 out of 5 patients had [ 18 F]CTT1057 positive lesions corresponding to pathology-proven PCa, while one patient had no focal prostatic uptake. In a second cohort suffering from metastatic castration-resistant PCa, the overall detection rate was higher with the novel compound compared to conventional imaging modalities, including bone scan and CT (Figure 4). Again, a late imaging protocol (90 min p.i.) increased the overall detection rate. In this cohort, all patients had definitive local therapy and the lowest PSA level was 0.7 ng/ml. Thus, the detection rate may be comparable to clinically established agents, such [ 18 F]PSMA-1007, and [ 18 F]DCFPyL [32,53,68], but future trials will investigate the accuracy of this agent in sites of disease, in particular in patients with low PSA levels [32]. LNCaP tumors, low accumulation in the skeleton, almost exclusive excretion from the kidneys and moderate to low liver uptake [77].
Clinical Investigations. In light of these encouraging findings, [ 18 F]FSU-880 was tested in six PCa patients with known metastatic disease. Radiotracer accumulation was noted in all sites of disease in the majority of the patients (5/6, with one patient under abiraterone acetate showing no uptake). No adverse events were recorded. The physiological biodistribution was similar to those of other small-molecule PSMA-targeting probes, with the kidneys receiving the highest dose, followed by the liver. Notably, four subsequent scans have been performed after injection of the radiotracer, with the final scan after 2 h. Again, a time-dependent increase in uptake was noted, with higher SUV max and tumor-to-blood ratios in the primary and metastatic lesions at later imaging time-points [33]. However, all of the included subjects had a Gleason score ≥7 and increased PSA-levels of 11.12 -487 ng/ml. Thus, further studies are needed to investigate the performance of [ 18 F]FSU-880 in patients with low PSA levels, preferably in a head-to-head comparison with the clinically established agents [ 18 F]PSMA-1007 and [ 18 F]DCFPyL [33]. Figure 5 shows such serially performed PET studies with [ 18 F]FSU-880 in a patient with bone involvement.

[ 18 F]JK-PSMA-7
Radiochemistry. In brief, [ 18 F]JK-PSMA-7 was prepared using a two-step reaction: In the first step, the radiolabeled active ester was produced by the nucleophile reaction of 18  high radiochemical yield, up to 40%, and in a high radiochemical purity (> 95%); the specific activity was 75 -120 GBq/μmol. The detailed procedure for the radiosynthesis using the "minimalist" protocol is described elsewhere [34].
Preclinical Investigations. As alluded to earlier, [ 18  was evaluated in peripheral ganglia of rats. As an underlying rationale, human ganglia are known to accumulate PSMA-targeted imaging agents and this model mimics small PSMA-expressing lesions comparable to the size of small lymph nodes (diameter of up to 3.5 mm) [34,78,79]. Whereas [ 18 F]JK-PSMA-7 showed similar image resolution relative to urea-based 18 F-labeled radiotracers, it demonstrated increased PSMA-specific cellular uptake as well as considerably higher imaging acutance as compared to [ 18 F]DCFPyL. Furthermore, [ 18 F]JK-PSMA-7 demonstrated higher target-tobackground ratios as compared to all other radiotracers [34].
Clinical Investigations. In vivo data has been acquired in a first-in-human study as well as in a biodistribution and dosimetry study (n=10) [80], in a direct comparison with [ 68 Ga]PSMA-11 (n=10) and in clinical applications in a larger clinical population (n=124, Figure 6) [81]. These studies demonstrate physiologic radiotracer accumulation in a pattern resembling the distribution known from other PSMA-targeted radiotracers with excretion via urinary and biliary pathways. Regarding dosimetry, whole body doses similar to other radiotracers were reported, with a maximum in the kidneys. High uptake in suspicious lesions was found, increasing over time, suggesting benefits of a late start of the PET/CT acquisition [80]. A pilot study in 10 patients who were examined with both PSMA-targeted radiotracers demonstrated that [ 18 F]JK-PSMA-7 was at least equivalent to [ 68 Ga]PSMA-11, as all [ 68 Ga]PSMA-11-positive lesions could be seen with [ 18 F]JK-PSMA-7 and several additional lesions were detected. In a subsequent analysis of a larger clinical cohort, [ 18 F]JK-PSMA-7 was found useful in various diagnostic scenarios (initial staging, biochemical recurrence and therapy monitoring). Notably, in patients with biochemical recurrence after prostatectomy (PSA level, ≥ 0.17 ng/ml) or in patients with biochemical recurrence after radiation therapy (PSA levels, ≥ 2 ng/ml), PSMA-positive lesions were detected in 44/53 patients (83%) [81]. Direct comparison with other 18 F-labeled PSMA-targeted radiotracers appears warranted, in particular as sensitivity for [ 18 F]DCFPyL has been described moderately lower in a similar clinical scenario (74.2%) [81].

[ 18 F]AlF-PSMA-11
Radiochemistry and Preclinical Investigation. Using a direct labeling procedure of PSMA ligands via aluminum fluoride 18 F-AlF-complexation, the novel imaging agent [ 18 F]AlF-PSMA-11 has been compared to [ 68 Ga]PSMA-11 in BALB/c nude mice with PSMA-expressing tumors. While tumor lesions could be appreciated with comparable uptake, [ 18 F]AlF-PSMA-11 demonstrated lower renal accumulation, which in turn renders this novel 18 F-labeled agent as an attractive compound for detection of small lesions close to the urinary tract. However, as a drawback, [ 18 F]AlF-PSMA-11 demonstrated a time-dependent increase of radiotracer uptake in the bone, and such defluorination may influence the accuracy of lesion detection in the skeleton [35]. The favorable binding affinities to PCa has also been proven by investigating PSMA-high LNCaP vs. PSMA-low PC3 tumors, showing a 24-fold higher uptake for [ 18 F]AlF-PSMA-11 in the high LNCaP tumors in C57BL6 mice [82]. Notably, a recently established automated synthesis of this compound will also allow for large scale production and gurantaee a high throughput of patients in a busy PET practice [83][84][85].
Clinical Investigations. In a recent dosimetry study, Piron and coworkers reported on a high safety profile in six patients with PCa. In addition, [ 18 F]AlF-PSMA-11 has a considerable low mean total radiation dose (comparable to [ 18 F]DCFPyL) [86]. Besides, a recent head-to-head comparison in 15 patients afflicted with PCa using both [ 68 Ga]PSMA-HBED-CC and [ 18 F]AlF-PSMA-11 showed that in additional 22% of the cases, bone lesions were only discernible with the latter compound [87].

Choosing the appropriate 18 F-labeled radiotracer
As demonstrated by the herein reviewed large variety of 18 F-labeled radiotracers, recent years have witnessed a tough competition in PSMA inhibitor imaging, which may be mainly driven by the ease of radiochemistry and the high incidence of PCa compared to other gender-specific tumor entities. Notably, PCa is the most common diagnosed cancer in men, while ovarian cancer ranks 7 th of the most commonly diagnosed tumor among women [88]. Given the wide array of different 18 F-labeled compounds for PCa imaging in men, it may be challenging to decide which radiotracer to introduce in one's institution. As outlined above, all of these imaging agents have demonstrated good imaging performance, potentially outperforming current diagnostic tools, e.g. Na 18 F PET, 68 Ga-PSMA PET/CT, CT, and/or bone scan. However, distinct radiotracer characteristics were noted among these 18 F-labeled radiotracers. [ 18 F]DCFPyL has very low hepatic uptake, which allows for the detection of small liver lesions even when cross-sectional correlates cannot be appreciated [14] or may be of value in later stages of disease [51]. [ 18 F]PSMA-1007 and [ 18 F]AlF-PSMA-11 have very low radiotracer accumulation in the urinary system, which renders these imaging agents an attractive alternative to identify small lesions in the pelvis or for local recurrence [35,69]. [ 18 F]JK-PSMA-7 may have higher target-to-background ratios and imaging acutance compared to [ 68 Ga]PSMA compounds as suggested by pre-clinical and clinical data [34,81]. [ 18 F]CTT1057 has a phosphoramidate core which may allow for irreversible binding to sites of disease and its theranostic twin has already been tested in a preclinical setting [32,75]. Nonetheless, the overall lesion detectability may be substiantially hampered by the different binding affinites and thus, IC50 values may serve as a reliable metric of which agent might generate the necessary radioactivity concentration to identify small volume sites of disease. For instance, [ 18 [74,77,89]. Thus, the affinity of the latter agent would be theoretically 30 fold higher compared to [ 18 F]DCFPyL (or 10 fold higher compared to [ 18 F]PSMA-1007). However, human PSMA-targeted PET imaging is a complex interplay of varying factors including biodistribution, radiotracer accumulation in putative sites of disease, renal excretion or normal variant uptake in benign lesions [90]. Thus, it is a matter of debate whether IC50 values derived by cell assays indeed reflect clinical reality [74]. In this regard, SUV may serve as a useful tool in clinical practice. Giesel  F]JK-PSMA-7 has also been found to be in the same range (12.8 ± 7.7) [32,80]. For the latter agent, the tumor-to-background ratios (with the gluteus muscle serving as reference) were significantly elevated, in particular in delayed scans [80] and such ratios would allow for a better comparison of the performance among all available radiotracers. However, considering the varying biodistribution of these agents, a semi-quantitative assessment of the tumor-to-bloodpool ratio, preferably compared to the net influx rate, would be desirable [91]. Table 3 highlights the advantages and limitations of the reviewed 18 F-labeled compounds.
In light of these distinct characteristics among different 18 F-labeled PSMA PET agents, an argument could be made that in the era of precision medicine, the use of the right radiotracer depends on what is currently needed by the interpreting nuclear medicine specialist or referring treating physician. Practically speaking, however, it is unlikely that all of these agents will become widely available or that they will all be available at a given site. However, that should not discourage users of other agents to also pursue the necessary multi-center prospective data to submit New Drug Applications for other 18 F-labeled PSMA-targeted compounds. Regulatory approval should also be sought in other jurisdictions. Ultimately, despite the nuanced differences in these compounds, all of them provide high diagnostic yield and the most-used compound will be that which is most widely available and conveniently accessed for a reasonable cost. Nonetheless, additional comparative studies are needed to firmly establish the role of all available 18 F-labeled compounds in the field of molecular PCa imaging, preferably in different clinical scenarios, which will allow for determining the true clinical utiltiy and benefit for patients [92]. Albeit first results have been recently published [51], such head-to-head comparisons on a larger scale are urgently needed. If not, we have nothing more than an embarrassment of riches, or in this case, radiotracers [92].  [14,31] Very low hepatic uptake, which allows for the detection of small liver lesions [14] May be of value in later stages of disease [51] One of the most extensively validated of the 18 F-labeled radiotracers for PSMA PET imaging Most likely the first of the herein reviewed compounds that will be approved by the U.S.

Food and Drug Administration [ 18 F]PSMA-1007
Higher hepatic background, which may be a drawback in later stages of disease (for the detection of liver lesions) [51] Very low radiotracer accumulation in the urinary system, which renders this imaging agent an attractive alternative to identify small lesions in the pelvis or for local recurrence [69] [ 18  Very low radiotracer accumulation in the urinary system, i.e. useful to identify small lesions in the pelvis recurrence [35] Future Perspectives high urinary excretion and, thus, these agents may benefit from modified imaging protocols to improve overall detection rate. These strategies include: intravenous application of furosemide, oral hydration, or voiding prior to imaging [32]. All 18 F-labeled agents have shown better target-to-background ratios in delayed scans and thus, future protocols should not start imaging acquisition prior to 90 min after radiotracer administration [31,32,52]. Moreover, as suggested by Schmuck et al. using [ 68 Ga]PSMA I&T, an early dynamic followed by a static delayed image acquisition also leads to an improved tumor-tonontumor ratio, in particular in the prostate gland [93] and this may be relevant for 18 F-labeled PSMA PET agents as well.
Risk Stratification and Application to Theranostics. For potential use in a theranostic setting, only [ 177 Lu]CTT1403, which is based on the same PSMA-binding scaffold as the PET agent [ 18 F]CTT1057, has been investigated in a preclinical environment [75]. Nonetheless, given the success of RLT based on [ 68 Ga]PSMA PET-based imaging, one may expect that 18 F-targeting PSMA agents will also be used to assess treatment rationale for a therapeutic approach with alpha-or beta-labeled PSMA ligands [12]. The field of PSMA-targeted theranostics has been fueled by the prolonged progression-free and overall survival in midgut neuroendocrine tumors treated with 177 Lu-based somatostatin receptor agonists [8]. However, as recently outlined by Bodei and coworkers, strategies to predict efficacy should be implemented in the clinic, otherwise nuclear medicine physicians may have to deal with a decline in interest by the referring oncologists, in particular as some patients may suffer from progressive disease under treatment [94]. In patients undergoing RLT, up to 20% do not respond with PSA decline and thus, future efforts should turn towards selecting patients most likely to benefit from such an endoradiotherapy [95]. Given the tentatively superior diagnostic performance of 18 F-labeled PSMA-targeted PET agents [18,19], risk stratification assessments based on these radiotracers may pave the way for a more reliable detection of treatment responders. However, prior to an application in a theranostic setting, 18 F-labeled PET-based findings should be further validated using histopathological reference, which has already been demonstrated for [ 68 Ga]PSMA-11 PET [96,97].
Other (pre)clinical competitors in the field of PSMA-directed imaging and treatment. Apart from the reviewed 18 [98]. 89 Zr-labeled monoclonal antibodies specifically designed for immuno PET imaging of PSMA expression have been also investigated in the preclinical and early clinical settings [99][100][101]. Recently, the generator produced radiopharmaceutical [ 44 Sc]Sc-PSMA-617 has been investigated and its long half-life of 4.04h renders it as an attractive PET agent for peri-therapeutic dosimetry studies [102]. Nonetheless, single photon emission computed tomography (SPECT) is more widely available as PET and PSMA-targeted SPECT compounds, such as the 111 In-labeled PSMA I&T, are a useful substitute if PET is not available [103]. Moreover, this compound has already proven its value in radioguided surgical procedures leading to high intraoperative detection rates of metastatic PCa lesions [104].

Conclusions
Due to inherent advantages, recent years have witnessed a shift from 68 Ga-labeled PSMA-targeted compounds towards 18  All of these 18 F-labeled PSMA PET imaging agents have demonstrated good imaging quality, potentially outperforming current imaging modalities. However, further research is warranted, e.g. to elaborate on the benefit of late imaging time-points or improved imaging protocols (e.g. by intravenous application of furosemide).