The main sites of prostatic skeletal metastases are ribs, thoracic and lumbar vertebrae, and ilium of the pelvis. Other less common sites of involvement are ischium, sacrococcyx, femur, skull, cervical vertebrae, sternum, scapula, and upper limb bones [21]. The bone involvement is commonly detected by BS and standard radiographs. Other imaging modalities such as CT, MRI, PET, and PET/CT can also be used in specific circumstances based on imaging timing, costs, radiation dose, and availability [22]. The current sequential approach is BS, BS plus TXR, and addition of CT or MRI in equivocal or discrepant cases [23].

¹⁸F-Sodium Fluoride (NaF) is gaining renewed interest and wider application for detection of osseous metastatic disease because of the widespread availability and improved quality of PET/CT scanners, but also because of the intrinsically higher spatial resolution of PET (compared to planar and SPECT imaging) and potential for PET quantitation of metastatic tumor burden [24]. NaF is a marker of bone perfusion and bone turnover, in which ¹⁸F-Fluoride ions exchange with hydroxyl groups in the hydroxyapatite crystal of bone to form fluorapatite with higher uptake in new bone (osteoid) because of higher availability of binding sites [24]. NaF PET/CT has been reported to be a highly sensitive and specific modality for detection of bone metastases in prostate cancer compared to standard ^99mTc-MDP (MDP) planar bone scans and MDP SPECT imaging [25]. The anatomic CT portion of the PET/CT was able to differentiate between benign and malignant lesions improving the specificity of NaF PET/CT vs NaF PET alone. Recent clinical practice guidelines for NaF PET/CT bone scans have been published by the Society of Nuclear Medicine and Molecular Imaging in 2010 [26]. A systemic review of the literature provides evidence for superior detection of bone metastases by both NaF PET and a functional tumor-based choline PET imaging agent (¹⁸F-fluorocholine or ¹¹C-choline), with or without CT, compared with conventional planar MDP bone scan [27]. This review reported a sensitivity and specificity for NaF PET or PET/CT of 88.6 and 90.7 %, respectively, on a per lesion basis, and 86.9 and 79.9 % on a per patient basis. It is noted that NaF uptake post-treatment flare phenomenon, similar to that observed with MDP bone scans, has also been observed, which should be taken into consideration [28]. Figure 4.2 shows examples of NaF PET/CT bone metastases and benign uptake in a degenerative vertebral osteophyte.

One of the benefits of NaF PET compared to ^99mTc-MDP (MDP) bone scans is the potential for quantitation of bone metastatic tumor burden and treatment response. A bone scan index using standard planar MDP bone scan had been proposed as an imaging biomarker to improve the reproducibility of treatment response assessment [29]. This index aimed to quantify tumor burden as a percentage of the total skeletal mass of a reference man but there are inherent limitations based on planar imaging. PET imaging with NaF for quantitative analysis for therapy response assessment would have potentially improved benefits over more limited planar and visual qualitative MDP bone scan-based assessment methods. One report in the literature using NaF PET for monitoring treatment response of bone metastases to radiotherapy ²²³Ra-Chloride was found to be more accurate than the visual qualitative assessment of scans, correlation with the serum-based PSA response or alkaline phosphatase activity [30]. A study of the kinetics and reproducibility of NaF PET/CT for detection and quantitation of bone metastases concluded an uptake period of 60 ± 30 min was sufficient for quantitation, with limited temporal dependence and demonstrated high tumor-to-normal tumor ratio [31]. Repeated baseline scans in this study also showed high intraclass correlations (>0.9) and relatively low critical percentage change (the value above which a change can be considered real) for these parameters, demonstrating relatively high reproducibility.

16β-¹⁸F-fluoro-5α-dihydrotestosterone (¹⁸F-FDHT) is an analog of dihydrotestosterone, the primary ligand for the androgen receptor (AR) which allows for non-invasive PET imaging of AR expression [49]. The first study of seven patients with progressive metastatic CRPC demonstrated ¹⁸F-FDHT detection of 78 % of the lesions compared to conventional imaging with average lesion SUV_max value of 5.28 [50]. Another study also demonstrated ¹⁸F-FDHT uptake at sites of metastatic disease, confirming AR-mediated ¹⁸F-FDHT PET signal as seen with interval decreased ¹⁸F-FDHT uptake from baseline (SUV_max decreased 9–70 %) after competition with flutamide [51].

The potential role of AR imaging as a pharmacodynamic marker in prostate cancer has been demonstrated in several recent studies. ¹⁸F-FDHT and ¹⁸F-FDG PET/CT imaging of a subset of 22 patients enrolled on a phase 1–2 study of MDV3100 (Enzalutamide) was able to show a reduction in ¹⁸F-FDHT tumor uptake from baseline (percent SUVmax change range from 20 to 100 %) after starting treatment compatible with competition or blocking of ¹⁸F-FDHT from the AR-binding site by the therapeutic agent [52]. However, these patients show a poor treatment response as only 45 % of these patients by ^18F-FDG PET/CT scans had 25 % or greater decrease in the FDG PET SUV_max after 12 weeks of therapy. These results position ¹⁸F-FDHT as a pharmacodynamic marker of AR binding rather than a therapy response biomarker. A more recent study also utilized ¹⁸F-FDHT-PET/CT imaging to measure pharmacodynamic response in a phase I clinical trial of a novel anti-androgen ARN-509 to incorporate imaging to visualize and quantitatively assess by SUV analysis the heterogeneity of AR-binding sites of metastatic disease and determine maximal AR inhibition by the study drug [53]. A clinically applicable method using ¹⁸F-FDHT PET quantitative marker as a surrogate of pharmacokinetic parameter to non-invasively assess free AR concentration has been proposed for validation studies in clinical trials [54].

An emerging and promising PET radiotracer for prostate cancer is anti-1-amino-3-^{18F-fluorocy-clobutane}-1-carboxylic acid (anti-¹⁸F-FACBC), a synthetic l-leucine amino acid analog. This agent is taken up in prostate cancer cells by amino acid transporters system (ASC transporter 2 or ASCT2) and to a lesser sodium-coupled neutral amino acid transporters but not incorporated into proteins intracellularly [55]. The initial first-in-human clinical study of anti-¹⁸F-FACBC PET/CT demonstrated positive uptake in primary prostate and metastatic disease [56]. Figure 4.3 demonstrates an example of anti-¹⁸F-FACBC PET/CT bone and nodal metastasis. In the subsequent clinical study, comparison of anti-¹⁸F-FACBC with ¹¹¹In-capromab pendetide (ProstaScint™) for detection of recurrent disease in 50 men after prostatectomy or external-beam radiation therapy for prostate carcinoma revealed anti-¹⁸F-FACBC PET/CT was more sensitive than ¹¹¹In-capromab pendetide SPECT/CT for detection of recurrent disease, especially for detection of extraprostatic recurrence. For disease detection in the prostate bed, anti-¹⁸F-FACBC had a sensitivity of 89 % (95 % CI: 74–97 %), specificity of 67 % (95 % CI: 35–90 %), and accuracy of 83 % (95 % CI: 70–93 %). For the detection of extraprostatic recurrence, anti-¹⁸F-FACBC had a sensitivity of 100 % (95 % CI: 69–100 %), specificity of 100 % (95 % CI: 59–100 %), and accuracy of 100 % (95 % CI: 80–100 %). A recent small study compared anti-¹⁸F-FACBC with 11C-Choline PET/CT imaging for detection of recurrent disease in 15 men after definitive therapy with prostatectomy or external-beam radiation therapy. Anti-18F-FACBC was found to have a higher detection rate compared to ¹¹C-Choline on a per patient (detection rate of 40 vs 20 %, respectively) and per lesion basis (6 vs 11 lesions, respectively), with all ¹¹C-Choline positive lesions also identified by anti-¹⁸F-FACBC [57]. These promising initial studies will need further validation of anti-18F-FACBC in larger multi-center clinical trials.

Prostate specific membrane antigen (PSMA) is a type II integral membrane protein expressed on the surface of prostate cancer cells, also known as glutamate carboxypeptidase II (GCPII) and folate hydrolase [58]. PSMA is associated with prostate cancer aggressiveness with histological studies associating high PSMA expression with metastatic spread [59–61], androgen-independence [62], and high PSMA expression levels predictive of prostate cancer progression [63, 64]. ¹¹¹In-capromab pendetide (ProstaScint™) was first developed for PSMA-based prostate cancer imaging but demonstrated limited performance for tumor detection which may be explained by lower tumor penetration of a large sized antibody agent and binding to the less accessible intracellular epitope of PSMA [65]. An improved humanized intact antibody that binds to the more accessible extracellular epitope of PSMA, J591, has been used labeled Indium-111 and Lutetium-177 for gamma planar and SPECT/CT imaging of metastatic prostate cancer as part of a number of radioimmunotherapy trials [66–70]. A recent retrospective review of these trials presented as a research abstract ranging over a decade demonstrated J591 planar imaging reported a detection rate of 86.4 % of known lesions [67]. Next generation imaging trials are incorporating PET/CT imaging of prostate cancer with ⁸⁹Zr-DFO-J591 which demonstrated high prostate tumor uptake in preclinical models [71].

Smaller low molecular weight imaging agents for PSMA would have inherent advantages over intact antibodies such as rapid tumor uptake and clearance from non-target sites [58]. N-[N-[(S)-1,3-Dicarboxypropyl]carbamoyl]-4-18F-fluorobenzyl-l-cysteine (¹⁸F-DCFBC) is a novel and clinically practical low molecular weight inhibitor of PSMA which in preclinical studies demonstrated high specific localization to PSMA-expressing prostate cancer xenografts [72]. An initial first-in-man study of ¹⁸F-DCFBC PET/CT was able to demonstrate uptake at sites of bone and lymph node metastatic disease detected at two hours after injection, with ¹⁸F-DCFBC PET detecting more lesions than corresponding conventional imaging modalities (CT, bone scan) [73]. These promising findings are undergoing further validation in ongoing clinical trials evaluating ¹⁸F-DCFBC PET/CT in primary and metastatic prostate cancer. Figure 4.4 demonstrates examples of positive detection of metastases by ¹⁸F-DCFBC PET/CT in men with CRPC seen with widespread bone and another with a large left pelvis nodal metastasis. Another fluorine-18 labeled low molecular weight PSMA targeted PET radiopharmaceutical, BAY 1075553, has been published recently with preclinical validation [74].

Other radiolabeled PSMA-based radiotracers for PET imaging including a Gallium-68 labeled PSMA-based low molecular weight radiopharmaceuticals have also been radiosynthesized [75]. Afshar-Oromieh et al. reported their initial experience with PET/CT using Glu-NH-CO-NH-Lys-(Ahx)-[⁶⁸Ga(HBED-CC)] (⁶⁸Ga-PSMA) which also was able to detect prostate carcinoma relapses and metastases at high signal-to-background at 1 h after radiopharmaceutical administration [76]. Another recent study comparing ⁶⁸Ga-PSMA with ¹⁸F-Fluorocholine with biochemical relapse of prostate cancer in the same patients within a 10-day time window in 37 patients was able to detect more lesions and higher signal-to-background by ⁶⁸Ga-PSMA compared to ¹⁸F-Fluorocholine (78 vs 56, respectively; P = 0.04) [77].

PSMA-based radiotracers have also been developed for SPECT imaging, which have lower resolution compared to PET scans but are more widely available than PET. First-in-man study of two Iodine-123 labeled small molecule SPECT agents for PSMA, ¹²³I-MIP-1072, and ^{123I-MIP-1095} also demonstrated uptake and detection of prostate cancer in soft tissue, bone, and the prostate gland as early as 1–4 h after injection [78]. Other SPECT agents for PSMA have been radiosynthesized with technetium-99 m demonstrating high specific PSMA targeted tumor uptake in preclinical models [79–81].

An interest and potentially powerful application of PSMA imaging is as a downstream marker of androgen-receptor signaling to potentially assess for real-time assessment of tumor response to androgen deprivation therapy. Evans et al. reported in preclinical studies that PSMA can be repressed by androgen treatment in multiple animal models of AR-positive prostate cancer in an AR-dependent manner, whereas anti-androgens can up-regulate PSMA expression [82]. In another study, assessment of androgen-receptor signaling in circulating tumor cells was able to utilize PSMA and PSA expression as markers of hormonally responsive prostate cancer to androgen deprivation therapy [83]. With the emergence of PSMA-based radiopharmaceuticals the possibility of utilizing non-invasive PSMA imaging by PET/CT or SPECT/CT as a downstream biomarker of androgen-receptor signaling may be an exciting possibility but requires careful clinical validation.