Preliminary Findings of the Role of FAPi in Prostate Cancer Theranostics

Prostate cancer (PCa) is the most frequently diagnosed cancer worldwide and the second most common cause of cancer-related deaths among men. Progress in molecular imaging has magnified its clinical management; however, an unmet clinical need involves the identification of new imaging biomarkers that complement the gold standard of prostate-specific membrane antigen (PSMA) positron emission tomography (PET) in cases of clinically significant PCa that do not express PSMA. Fibroblast activation protein (FAP) is a type II transmembrane serine overexpressed in many solid cancers that can be imaged through quinoline-based PET tracers derived from an FAP inhibitor (FAPi). Preliminary results of FAPi application in PCa (in PSMA-negative lesions, and in comparison with fluorodeoxyglucose—FDG) are now available in the literature. FAP-targeting ligands for PCa are not limited to detection, but could also include therapeutic applications. In this preliminary review, we provide an overview of the clinical applications of FAPi ligands in PCa, summarising the main results and highlighting contemporary strengths and limitations.


Introduction
Prostate cancer (PCa) is the most frequently diagnosed cancer worldwide and the second most common cause of cancer-related deaths among men [1]. Prostate-specific antigen (PSA) is the most useful biochemical parameter in PCa clinical practice [2] with a more limited value at the advanced stage due to increased PCa heterogeneity and de-differentiated lesions, also considering that optimal intervals for PSA follow-up are unknown [3]. Fortunately, the progress of next-generation imaging in the past few years has revolutionised PCa management. For example, using tracers that target the prostate-specific membrane antigen (PSMA) in positron emission tomography (PET) has become a well-established technique for the assessment of biochemical recurrence (BCR) [4], as well as further applications for biopsy guidance [5], staging [6] and the evaluation of disease progression [7]. However, the diagnostic accuracy of PSMA PET is challenged by findings that claim that up to 15% of clinically relevant PCa cases do not express this protein [8], leading to false negative results. Moreover, further cancer de-differentiation in the metastatic castration-resistant (mCR) late stages of the disease increases this percentage [9,10]. Therefore, identifying alternative imaging biomarkers that complement PSMA PET in such clinical circumstances represents an unmet clinical need.
In the last few years, the tumour microenvironment (TME) has gained attention due to its role in tumourigenesis, neo-angiogenesis and cancer progression. The TME consists of immune cells, extracellular matrix, vasculature, and cancer-associated fibroblasts (CAFs), which are the primary stromal cells within the TME. They can be identified based on the expression of several markers [11], including the fibroblast activation protein (FAP), which can be targeted by quinoline-based PET tracers derived from an FAP inhibitor (FAPi) [12]. FAP is a type II transmembrane serine protease that is nearly absent in healthy tissues, benign tumours, and normal fibroblasts (except in cases of chronic inflammatory conditions). However, in over 90% of the most common human epithelial tumours, it is overexpressed in the stroma and is histologically characterised by desmoplastic reactions [12]. FAP overexpression increases the risk of tumour invasion, lymph node metastasis and the decreased overall survival (OS) [12][13][14][15] of many solid cancers, including mCRPCa [16], especially if the cancer exhibits neuroendocrine (NE) differentiation [17]. It thus represents a potential target for theranostics applications for aggressive (and often, low-PSMA-expressing) PCa [13,[18][19][20][21][22][23][24][25]. In this systematic review, we provide an overview of clinical applications using FAPi ligands in the PCa theranostics scenario by summarising the main preliminary results.

Materials and Methods
A literature search was performed by six independent authors (R.L., V.D., M.L., A.N., A.R. and G.C.) using the PubMed, Scopus, Google Scholar, Cochrane and EMBASE databases regarding the role of FAPi PET in PCa. The following terms were used: "FAP" or "fibroblast activation protein" or "FAPi" and "prostate cancer" or "prostate" and "PET" or "positron emission tomography" or "RLT" or "PRRT". Papers (including original articles, clinical studies, and case reports of humans) that were written in the English language and had been published before September 2022 were included in the review. Editorials and reviews were excluded. Duplicate papers were removed. All papers were screened by title and abstract content. Full texts of papers that were inherent to the endpoint of the present review were retrieved to verify their relevance. Moreover, to enrich the collected data, all the references in the selected papers were also checked. The following data were extracted from the selected papers: name of the first author, year of publication, type of FAPi tracer used, sample size, number of PCa, the scenario, International Society of Urological Pathology (ISUP), prostate-specific antigen (PSA) values (when these were available) and the main findings.

Results
The literature search returned 65 studies. Five duplicate papers were eliminated. After screening the titles and abstracts, 36 full-text papers were retrieved. Of these, 21 articles did not fulfil the selection criteria and were excluded. Figure 1 details the selection process workflow according to the PRISMA guidelines [26].
The 15 remaining articles were published by researchers from Europe and Asia whose studies were conducted on 1-6 patients with PCa who underwent FAPi PET or FAPi radioligand therapy (RLT). Tables 1-4 describe the studies' main characteristics and results according to the considered PCa scenarios. The 15 remaining articles were published by researchers from Europe and Asia whose studies were conducted on 1-6 patients with PCa who underwent FAPi PET or FAPi radioligand therapy (RLT). Tables 1-4 describe the studies' main characteristics and results according to the considered PCa scenarios.

FAPi Immunohistochemistry (IHC) and Biodistribution Studies of PCa
Kesch et al. [11] aimed to evaluate FAP expression in tissue microarrays (TMAs, quantified through the H-index) from radical prostatectomy (RPE) or transurethral prostate resection (TURP) specimens of prostatic tissue from patients at different stages. They reported mean H-indexes for benign tissue of 0.018, for primary PCa of 0.031, for neoadjuvant androgen deprivation therapy (ADT) before RPE of 0.042, for CRPCa of 0.076 and for NEPCa of 0.051, indicating a significant rise in FAP expression as the disease progressed, as already described in a preclinical model [16]. They also imaged two patients with progressive mCRPCa and one with NEPCa and found [ 68 Ga]Ga-DOTA-FAPi-04 describing multiple metastatic sites with high uptake, which confirmed their array's results.
A recent study by Greifenstein et al. [13] assessed the "kit-type" synthesis of the CAFtargeting tracer [ 68 Ga]Ga-DATA5m.SA.FAPi and analysed the tracer's uptake and biodistribution in one patient with sarcoidosis and five patients with various histologically confirmed metastatic malignancies, including one patient with PCa. With a very low physiological uptake in the brain, liver, intestine, bone and lungs, [ 68 Ga]Ga-DATA5m.SA.FAPi showed a biodistribution similar to that of FAPi-02, but with less kidney uptake and blood pool activity and a higher uptake of [ 68 Ga]Ga-DATA5m.SA.FAPi than FAPI-46 and FAPi-

FAPi Immunohistochemistry (IHC) and Biodistribution Studies of PCa
Kesch et al. [11] aimed to evaluate FAP expression in tissue microarrays (TMAs, quantified through the H-index) from radical prostatectomy (RPE) or transurethral prostate resection (TURP) specimens of prostatic tissue from patients at different stages. They reported mean H-indexes for benign tissue of 0.018, for primary PCa of 0.031, for neoadjuvant androgen deprivation therapy (ADT) before RPE of 0.042, for CRPCa of 0.076 and for NEPCa of 0.051, indicating a significant rise in FAP expression as the disease progressed, as already described in a preclinical model [16]. They also imaged two patients with progressive mCRPCa and one with NEPCa and found [ 68 Ga]Ga-DOTA-FAPi-04 describing multiple metastatic sites with high uptake, which confirmed their array's results.
In a basket study, Giesel et al. assessed the biodistribution and preliminary dosimetry of [ 68 Ga]Ga-DOTA-FAPi-02 and [ 68 Ga]Ga-DOTA-FAPi-04 in 50 patients with various cancers including four cases of PCa using FDG as the standard of reference. A favourable biodistribution was found for both tracers, with fast tracer kinetics (prolonged retention for [ 68 Ga]Ga-DOTA-FAPi-04). Both tracers were unaffected by the blood sugar levels and had similar dosimetry to other 68Ga compounds and [ 18 F]FDG, enabling potential theranostics applications owing to the DOTA chelator. No specific data were reported for the patients with PCa [12].
A recent study by Greifenstein et al. [13] assessed the "kit-type" synthesis of the CAF-targeting tracer [ 68 Ga]Ga-DATA5m.SA.FAPi and analysed the tracer's uptake and biodistribution in one patient with sarcoidosis and five patients with various histologically confirmed metastatic malignancies, including one patient with PCa. With a very low physiological uptake in the brain, liver, intestine, bone and lungs, [ 68 Ga]Ga-DATA5m.SA.FAPi showed a biodistribution similar to that of FAPi-02, but with less kidney uptake and blood pool activity and a higher uptake of [ 68 Ga]Ga-DATA5m.SA.FAPi than FAPI-46 and FAPi-74 in the malignant lesions. The patient with PCa had metastatic small cell extensive disease (ISUP 3) and the highest [ 68 Ga]Ga-DATA5m.SA.FAPi maximum standardised uptake value (SUV max 13.7) in bone metastasis among the whole cohort.
Kratochwil et al. performed a retrospective analysis [19] to quantify the uptake of [ 68 Ga]Ga-DOTA-FAPi-04 in 80 patients with 28 different tumour entities (confirmed by histopathological or unequivocal radiologic findings), including 4 PSMA-negative PCa cases. The authors defined low, intermediate and high uptake using SUV max cut-offs of <6, from 6 to 12, and >12, respectively. In one PSMA-negative PCa case, the authors observed a high-intermediate [ 68 Ga]Ga-DOTA-FAPi-04 uptake; however, in three NEPCa cases, the authors reported low uptake (SUV max < 6), which is different from the findings of Vlachostergios et al., who reported high mRNA FAP expression levels in 21 patients with NEPCa who received previous treatment with taxanes and abiraterone or enzalutamide [17]. The main characteristics and results of the studies described in this paragraph are reported in Table 1.  [6][7][8][9][10][11][12] was observed in a patient with PSMA-negative PCa.
Low uptake (SUV max < 6) was observed in 3 patients with NEPCa.

Comparison of FAPi and PSMA PET for PCa
The aim of the study by Kessel et al. [18] was to compare FAP and PSMA expression levels in primary PCa by performing histological analyses and PET imaging in a small cohort of 11 PCa patients with Gleason scores of >7 (5 staging and 6 mCRPCa) and 3 patients with benign prostatic hyperplasia. Among the patients, the 6 patients with mCRPCa underwent both [ 18 F]PSMA-1007 and [ 68 Ga]Ga-DOTA-FAPi-46 PET/computed tomography (CT). The authors observed that the IHC FAPi expression level in the patients with prostatic disease was not necessarily found only in those with CAF, but also in those with benign myofibroblasts in the context of suspected chronic prostatitis. Nevertheless, The technique of [ 68 Ga]Ga-DOTA-FAPi-46 PET/CT was useful for an advanced disease, as confirmed in a patient with NEPCa who had more FAPi-positive lesions than they did PSMA-positive lesions. This preliminary finding hampers the diagnostic relevance of FAPi in early PCa, as it is unspecific especially in the setting of prostatic inflammation.

Comparison of FAPi and FDG PET for PCa
In an international retrospective multicentre analysis, Giesel et al. [20]  In a case report, Xu T et al. [28] described the findings of [ 68 Ga]Ga-DOTA-FAPi-04 and [ 18 F]FDG PET/CT in a 76-year-old man with a history of chronic prostatitis, left shoulder osteoarthritis and a magnetic resonance imaging (MRI)-detected nodule in the left frontotemporal lobe, which was suspicious for brain metastasis. With a moderately elevated PSA level of 4.6 ng/mL, the authors observed abnormal uptake with both tracers in the known intracranial lesion, in the left shoulder and in the left peripheral zone of the prostate. A biopsy of the left posterolateral prostate region confirmed the presence of an ISUP 2 PCa. The patient also showed more tracer uptake at the site of the shoulder arthritis than that at the tumour site, indicating the potential of usefulness of [ 68 Ga]Ga-DOTA-FAPi-04 imaging in cases with inflammation. In this case report, the increased [ 68 Ga]Ga-DOTA-FAPi-04 uptake was similar to that shown by [ 18 F]FDG imaging, which suggests that [ 68 Ga]Ga-DOTA-FAPi-04 imaging may not be more tumour-specific than [ 18 F]FDG is for PCa. The main characteristics and results of the studies described in this paragraph are reported in Table 3.

FAPi Theranostics Applications for PCa
In a case report, Khreish et al. [25]  uptakes of all metastases (higher than those of PSMA and FDG), which suggests that in PCa, FAP-targeted radionuclide therapy (using molecules with longer retention times than those of FAPi-04) has the potential to overcome limitations related to tumour heterogeneity and an insufficient PSMA expression level.

Conclusions
Representing a wide spectrum of probes, [ 68 Ga]Ga-DOTA-FAPi ligands have certain advantages for cancer assessment, including a high TBR, rapid renal clearance, fast tracer kinetics, no blood glucose level dependency and no need for resting.
It is well known that PCa is a highly heterogeneous disease, in which molecular imaging endorses a fundamental role. However, the characteristics of patients with low-PSMA-expressing PCa who may benefit from new probes are a "grey area" [31]. Stateof-the-art literature regarding FAPi ligands in PCa is scarce, and this warrants further investigation, as only a few studies have assessed the clinical role of FAPi PET or FAPi RLT in the PCa scenario, with each cohort including only a maximum of 6 patients (for a total of 32 patients), providing scarce specific data [16][17][18]21]. In PCa theranostics, particular attention must be paid to the selection/use of FAPi ligands in terms of the retention time. Among the FAPi molecules that have been explored so far in PCa, FAPi-46 showed a longer retention time than FAPi-04 did, making it a more suitable ligand for RLT. The authors showed the feasibility of FAPi theranostics using a beta probe ([ 90 Y]DOTA-FAPi-46) [24] and a probe with a longer half-life ([ 177 Lu]Lu-FAPi-46) [23].
On the basis of our preliminary systematic review, we summarise that FAPi PET may potentially improve the diagnostic accuracy of next-generation imaging in patients with PCa with heterogeneous PSMA expression levels or at later stages of the disease. The role of FAPi PET for NEPCa is still controversial. Many issues remain to be solved, including enlarging the sample sizes of clinical studies, designing prospective studies with accurate methodologies to determine the exact diagnostic performance of FAPi PET in all phases of the disease and identifying ligands with a longer retention time in tumour tissues.

Data Availability Statement:
The data analysed during this review study are available from the corresponding author upon reasonable request.

Conflicts of Interest:
The authors have no relevant financial or non-financial interest to disclose.