Clinical advancement of precision theranostics in prostate cancer.

Prostate cancer (PCa) is the most frequently occurring cancer in men worldwide, with a continuously increasing incidence ( 1 ). Traditional methods for PCa diagnosis, including the digital rectal examination (DRE) and serum prostate-specific antigen (PSA) evaluation, cannot fully meet the diagnostic needs due to low accuracy and sensitivity ( 2 ). Novel methods, such as integrated positron emission tomography/computed tomography (PET/CT) or PET/magnetic resonance imaging (PET/MRI), to image 68 Gallium( 68 Ga)-labeled prostate-specific membrane antigen (PSMA), which is exclusively overexpressed on clinical PCa cells, have brought great precision diagnostic capability. In addition to diagnosis, the major treatment strategy for PCa, prostatectomy, has entered the era of “precision surgery”, which requires a precise marking of the malignant tissue as intraoperative guidance. Identifying the actual position of the tumor, nerve, and lymph node has become more and more important during prostatectomy surgery. Novel intraoperative molecular imaging methods with high sensitivity, specificity, distinguishability, and safety, such as 111 In labeled PSMA, have been shown to locate the PCa lesions precisely ( 3 ); indocyanine green (ICG), a USA Food and Drug Administration (FDA)-approved near-infrared (NIR) fluorescent agent for highlighting tissue, has been combined with 99m Tc to directly and accurately recognize malignant PCa tissue and metastases to assist decision making by surgeons during operations ( 4 ). In this article, we focus on providing a comprehensive summary of all novel molecular imaging probes in PCa diagnosis and intraoperative guidance for tumor detection and nerve preservation.

Compared to diagnostic imaging tracers, there are fewer tracers available for guidance during PCa surgery. Herein, we present some novel intraoperative tracers, which are promising methods for PCa precision surgery in the future. Indocyanine green (ICG), one of the most common near-infrared (NIR) fluorophores for fluorescence-guided surgery (FGS), has been approved by the FDA for more than 60 years. It is a 776 Da, amphiphilic tricarbocyanine, water-soluble, and anionic probe. It binds to protein quickly and is confined to the intravascular compartment through intravenous injection ( 76 ). The half-life of ICG is 150-180 seconds, and it has low toxicity. Glutathione S-transferase, a transport protein, is able to make ICG through the liver and excrete into bile totally; thus, ICG can be administered repeatedly every 15 minutes during surgery to label the tissue ( 77 ). Due to its relatively low cost and widespread availability, ICG is widely used in urologic surgery, including laparoscopic and robotic adrenalectomy procedures ( 78 , 79 ). In laparoscopic robot-assisted radical prostatectomy (RARP), Mangano et al. used ICG with NIR fluorescence to guide the preservation of the neurovascular bundle ( 80 ). Tobis et al. adopted ICG to highlight the renal vasculature and distinguish between normal and malignant tissue ( 81 ). Rho et al. used CT to guide the penetration of ICG through fluorescence thoracoscopy, precise location and margin resection of the radiopaque lesions were confirmed via C-arm fluoroscopy, and pulmonary nodules were resected with an endostapler ( 82 ). As a result, the ICG imaging guided pulmonary nodule removal was 100% in the 24 patients. However, due to the nature that ICG is a non-targeted probe with suboptimal emission characteristics for NIR-II detection, it cannot distinguish between benign and malignant tumors and can be accumulated by other tissues, which may cause false positives ( 83 ). This disadvantage was shown by Tummers et al. in a study on oncologic procedures of fluorescence-guided surgery with a high false-positive rate (62%) for the application of ICG ( 84 ). As mentioned, PSMA is a type II integral membrane glycoprotein that shows elevated expression in the majority of PCa cells ( 85 ). It is a marvelous target for image-based intraoperative guidance for accurate tumor identification due to three reasons. First, PSMA is exclusively overexpressed on tumor cells of primary PCa lesions, while its expression is consistently low in healthy prostate tissues. Second, the expression level of PSMA correlates with the Gleason grading of PCa lesions. Last, binding with the extracellular domain of PSMA normally induces internalization of the imaging agents, resulting in substantial retention of the labeling inside the tumor lesions ( 86 ). PSMA radio-guided surgery (PSMA-RGS) has been approved to be an efficient method for resecting primary tumors and metastatic lymph nodes ( 87 ). Intravenous injection of 111 In-labeled PSMA-I&T to PCa patients during surgery has enabled the visualization of metastatic lymph nodes, which are normally unobtrusive and unrecognizable ( 88 ). Clinically, in patients undergoing salvage lymphadenectomy, the 111 In-PSMA-RGS allows intraoperative detection of small lymph node metastases with high specificity and sensitivity ( 89 ). In addition, the 111 In-PSMA-617 tracer also helped surgeons deal with unidentified pelvic lymph node metastases in situ during the surgery and resected ex vivo tissue samples to prove the successful removal ( 90 ). Except for 111 In-labeled PSMA ligands for detecting metastases of PCa, Robu et al. explored another ligand named 99m Tc-mas3-y-nal-k(Sub-KuE) for PCa imaging ( 91 ). Clinically, 99m Tc is preferable to 111 In, as it provides low-energy gamma rays that are more suitable for RGS due to the high sensitivity of gamma probes for collimation, and 99m Tc has a much shorter half-life (6 hours) than 111 In (2.8 days), resulting in faster pharmacokinetics and lower radiation exposure for both patients and nuclear medical professionals ( 92 ). The hybrid tracer ICG- 99m Tc-nanocolloid combining fluorescent dye ICG with the radioactive 99m Tc-nanocolloid, not only offers preoperative sentinel node (SN) mapping, but also provides better optical surgical guidance ( 93 ). The tracer shows no leakage into the surgical field and provides a depth estimation (>0.5–1 cm) of the nodal location, which helps to prevent surgery-related side effects ( 94 ). Another study also approved the value of the hybrid tracer in the surgical identification of lymph nodes ( 95 ). Overall, one obvious advantage of the ICG- 99m Tc-nanocolloid tracer is that it can enable visualization of any tumor lesion or SN in their anatomical context during surgery, and its application is independent of the order of resection (primary tumor or metastasis) or the surgical setting (open or laparoscopic) ( 94 ). Iatrogenic nerve injuries are common in prostatectomy, 20% of postoperative patients suffer from urinary incontinence, and many patients experience erectile dysfunction, which can only be partially mitigated by existing nerve-sparing surgical techniques. It is challenging to intraoperatively identify the specific location of buried small peripheral nerves (PNs), but the endeavor to find new ways to protect PNs is significant ( 96 , 97 ). To meet the clinical need, an ideal method for imaging PNs during the intraoperative procedure should possess the following features. First, a high specificity and a good signal-to-noise ratio are essential. Second, real-time and long-term imaging is vital for PNs to be recognized and retained during surgery ( 98 ). Third, the imaging probes should have good biosafety. Last, the cost should be low enough for clinical use ( 97 ). Neurovascular dyes such as ICG and fluorescein have been used to highlight PNs in clinical settings ( 96 ). It has been shown that fluorescein was applied to visualize abnormal peroneal nerves in ganglion cyst excision procedures ( 99 ). Recently, ICG has been used to help protect critical functional structures in prostatectomy by enabling the identification of all neurovascular bundles without increasing the operative time or complications ( Figure 4 ) ( 80 ). These promising data indicate that iatrogenic injury can be prevented, and the operative time can be shortened with the help of fluorescence-guided imaging. According to the clinical study performed by Jin et al., in patients with ICG injected 24 hours prior to surgery, the pelvic autonomic nerves can be intraoperatively seen clearly under a NIR ray ( Figure 5 ) ( 100 ). Due to the ubiquity of such fluorophores, it is foreseeable that surgeons will attempt the fluorescent nerve-targeting agents more frequently in their clinical practice ( 96 ). However, the agents can have light penetrance through the tissue of greater than 5-6 mm. Such a deep penetration causes increased light scatter, thus obscuring the specific location of PNs. In addition, ICG is not a targeted dye, and it is not able to distinguish the nerve bundle from other tissues. For example, in the surgery for deep endometriosis, the ischemic lesion, the hypogastric nerve, the pelvic plexus, and the ureter were all dyed by ICG ( 101 ). Prostate vascularization and neurovascular bundles by ICG. Reprinted with permission from Mangano et al. ( 80 ). Copyright © 2017 Wichtig Publishing. Autonomic pelvic nerves under the fuorescence (A) and under the white light (B) , the sacral plexus of autonomic pelvic nerves are displayed very clearly under NIR ray (white arrows) but not clearly under white light. Reprinted with permission from Jin et al. ( 100 ). Copyright © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2021. Extensive studies have reported that PCa tumor progression is favored by innervation. Magnon et al. reported that the formation of autonomic nerve fibers in the prostate gland regulates the development and dissemination of PCa ( 102 ). Therefore, biomarkers for innervation and effective visualization methods are necessary to assess nerve density in PCa. Nerve peptide 41 (NP41) has been found as a marker to highlight peripheral nerve tissue, and fluorescent-labeled NP41 can be visualized through its binding to the motor and sensory nerves in live mice ( 103 , 104 ). Hingorani et al. reported that NP41 had the best nerve-to-non-nerve contrast compared to other peptides like NP38, 40, and 42, and the average nerve-to-non-nerve signal ratio increases by 17% under fluorescent imaging compared to white light ( 105 ). NP41 is considered an excellent agent for in vivo tracking of nerves in rodents. Since NP41 specifically targets nerves in PCa, it has the potential for visualizing nerve density and tumor innervation in PCa. The nanoprobes named propranolol-loaded-superparamagnetic iron oxide (SPIO)-NP41 nanoparticles (PSN NPs) have been used to assess the nerve density of PCa with high sensitivity and high specificity in mice ( 106 ). Since PSN NPs had an exclusive accumulation at the tumor site, benefiting the targeted delivery of propranolol, this study showed that PSN NPs inhibited PCa tumor growth by blocking the interaction between tumor cells and sympathetic nerves in the neural tumor microenvironment. Nevertheless, existing data on applying NP41 to ex vivo human nerve tissue provided little contrast compared to muscle ( 105 ). Hence, human NP401 (HNP401), a peptide that binds to and highlights human autonomic and motor/sensory nerves, was identified for improving the labeling of human nerves, especially for the human prostate gland, suggesting its potential guidance role in the prostatectomy for PCa patients.

Gastrin-releasing peptide receptor (GRPR) is a G protein-coupled receptor that is overexpressed in a variety of malignancies, such as breast cancer, PCa, and small cell lung cancer ( 65 ). GRPR is one of the subtypes of the bombesin (BBN) receptor, also called BB2r. As a bombesin analog, gastrin-releasing peptide (GRP) spreads over the peripheral nervous system and organs and primarily works in the gastrointestinal system through GRPR ( 66 ). As mentioned, the critical feature of GRPR is its overexpression in prostate tumor cells and underexpression in normal prostate tissue. Therefore, multiple radionuclides have been used to label bombesin analogs (GRPR agonists and antagonists), which preserve the high affinity for GRPR, to image tumors with high GRPR expressions ( 67 , 68 ). At present, a variety of GRPR agonists and antagonists have emerged and been tagged with multiple radioisotopes. However, the GRPR agonists induce some gastrointestinal side effects due to the activation of GRPR. Compared with agonists, GRPR antagonists could provide better visualization with high value in the diagnosis and staging of PCa with less undesirable effects ( 69 ). As one of the GRPR antagonists, RM26 was radiolabeled to trace the GRPR in prostate tumor tissues. In Zhang’s study, both NOTA-RM26 and agonist BBN were labeled with 68 Ga to image the lesions in 22 PCa patients ( 70 ). The results showed that the 68 Ga-RM26 tracer visualized much more primary lesions and metastases with significantly higher SUVmax than 68 Ga-BBN PET/CT ( Figure 3 ) . Bakker et al. performed 68 Ga-SB3 PET/CT imaging on 10 PCa patients before radical resection with a sensitivity of 88% and a specificity of 88% in 16 lesions detected by prostatectomy pathology, suggesting that 68 Ga-SB3 PET/CT could be used for the detection and localization of primary PCa ( 71 ). Duan et al. compared 68 Ga-RM2 PET imaging with multiparametric magnetic resonance imaging (mpMRI) and 68 Ga-PSMA-11 PET on 41 patients with the initial diagnosis of intermediate and high-risk PCa. 68 Ga-RM2 and 68 Ga-PSMA11 had similar sensitivity and accuracy of 98%, 89% and 95%, 89%, respectively, which are significantly higher than mpMRI with 77% and 77%, for the detection of intraprostatic lesions ( 72 ). The post-prostatectomy histopathology also affirmed the ability of 68 Ga-RM2 PET imaging with a detection rate of 93%. Comparison of 68 Ga-RM26 PET/CT (A) , and 68 Ga-BBN PET/CT (B) in a 73-y-old man diagnosed as having PCa (white arrow) with lymph node involvement (red arrow) and bone metastasis (yellow arrow) before prostatectomy. 68 Ga-RM26 PET/CT detected primary tumors, multiple lymph node involvement, and bone metastasis lesion, whereas those lesions showed much lower uptake on 68 Ga-BBN PET/CT. Reprinted with permission from Zhang et al. ( 70 ). Copyright © 2018 by the Society of Nuclear Medicine and Molecular Imaging. Not only for the initial diagnosis of PCa, but GRPR-targeted PET imaging could also take a role in the follow-up with the detection of BCR. Minamimoto et al. conducted a prospective study of 32 patients with BRC of PCa but negative imaging results on multiple conventional imaging modalities (CT, MRI, and 99m Tc-MDP bone scan) ( 73 ). Among the 32 participants, 23 individuals were recognized through the 68 Ga-RM2 PET imaging, suggesting a detection rate of 71.8% in these patients without positive findings on conventional imaging tools. Wieser et al. also collected 16 choline-PET/CT-negative/indeterminate biochemically recurrent PCa patients to evaluate the imaging ability of 68 Ga-RM2-PET/CT in detecting metastatic tumors and found that tumors in 10 out of 16 patients (62.5%) could be recognized by the 68 Ga-RM2-PET/CT imaging ( 74 ). In addition, the expression of GRPR appears to be unassociated with PSMA, suggesting that GRPR and PSMA-targeted PET imaging could be complementary ( 75 ). Therefore, GRPR-targeted imaging could complement other conventional modalities. Thus, GRPR tracer imaging is a promising tool for diagnosing and surveillance PCa with its high detection rate. However, the low sample size in these clinical trials critically affects the credibility of the evaluation of GRPR tracer imaging. More clinical trials with larger sample sizes are necessary for the future. In conclusion, various imaging agents for the precision diagnosis of primary and metastatic PCa are under study, with both advantages and disadvantages ( Table 1 ) . Although there is no 100% satisfactory imaging agent for PCa at present, with the in-depth research on current imaging agents and the development of new imaging agents, multi-target combined imaging or individualized imaging may bring better clinical value to PCa patients. Pros and cons of PET imaging agents for prostate cancer.

For tumor PET imaging, 18 F-fluorodeoxyglucose ( 18 F-FDG) is one of the most frequently used radiotracers. Fluorodeoxyglucose (FDG) is a glucose analogue, which is highly absorbed in tumor lesions mainly through glucose transporter-1 (GLUT1) because of its involvement in tumor cell metabolism. It has been broadly applicated in clinical diagnosis, staging analysis, prognosis prediction, and treatment response monitoring of various tumors as a PET imaging agent ( 45 ). However, some patients with well-differentiated PCa had false negatives during clinical imaging ( 46 ). In addition, some benign lesions, such as inflammation, can also take up 18 F-FDG. Since the prostate is close to the bladder and 18 F-FDG is mostly egested through the urinary tract, this limits its application in the primary tumor of PCa due to the bladder urinalysis activity ( 9 ). Although 18 F-FDG imaging possesses limited accuracy on primary PCa diagnosis and staging, high-grade PCa (GS= 8-10) and more aggressive metastatic PCa showed higher glycolytic activity. In a study of 148 PCa patients with biopsy GS ≥ 8, 18 F-FDG PET/CT imaging detected lesions with high intraprostatic FDG uptake in 66% of patients ( 47 ). Intraprostatic FDG uptake was positively correlated with higher pathological GS, seminal vesicle invasion, pathological lymph node metastasis, and risk of BCR, suggesting that preoperative intraprostatic FDG uptake is a composite factor for poor pathological prognostic factors. In addition, 18 F-FDG has a certain value in the detection of primary lesions of CRPC. Chen et al. studied 56 cases of CRPC with 68 Ga-PSMA and 18 F-FDG PET/CT examinations ( 48 ). Although overall the 68 Ga-PSMA is significantly better than 18 F-FDG PET/CT with a higher detection rate of 75.0% vs 51.8%, and more positive lesions of 135 vs 95, the incidence of patients with 68 Ga-PSMA−, 18 F-FDG+ lesions was 23.2% (13/56), which could not be ignored in the clinic. The PSA level and GS of patients with 68 Ga-PSMA−, 18 F-FDG+ lesions were higher than those of patients without 68 Ga-PSMA−, 18 F-FDG+ lesions, that 61.5% of patients with GS ≥ 8 and PSA ≥ 7.9 ng/mL carried the special lesions, suggesting that CRPC patients with high GS and PSA may take advantage of 18 F-FDG PET/CT imaging. 18 F-FDG-PET/CT is also of great value in the diagnosis of bone metastases in high-grade PCa patients (GS≥8). In comparison with the bone scan, 18 F-FDG PET/CT is sensitive and accurate in detecting bone metastases (sensitivity:100% vs 78.8%, specificity: 98.7% vs 98.2%) ( 49 ). 18 F-FDG PET imaging also has the ability to assess prognosis in PCa patients. In the study of 94 patients with primary PCa who underwent 18 F-FDG imaging previous to the radical prostatectomy, patients with higher SUVmax had poorer long-term survival ( 50 ). Higher intensity tracer uptake is positively associated with GLUT1 expression, stage, pathological grade, and disease progression. 18 F-FDG PET whole-body total lesion glycolysis (TLG) is independently associated with overall survival as a quantitative prognostic imaging biomarker in mCRPC patients receiving abiraterone or enzalutamide as first-line therapy ( 51 ). Studies have shown that SUV value and the number of lesions are also independently associated with time to hormonal therapy failure (THTF). When the sum of SUVs was divided into quartile ranges, patients in the fourth quartile had significantly lower odds of survival than patients in the first quartile. Both SUV and 18 F-FDG PET/CT-derived lesions provide independent prognostic information for THTF in patients with metastatic castration-sensitive PCa ( 52 ). FDA approved the application of choline-based radiotracers ( 11 C and 18 F- choline) in 2012 for patients with biochemically relapsed PCa. Now both 11 C and 18 F- choline have been applied to monitor the curative effect in PCa patients. Wang et al. analyzed 46 studies and found that the combined sensitivity and specificity of 18 F-choline for the detection of BCR of PCa were 0.93 (95% CI, 0.85-0.98) and 0.91 (95% CI, 0.73-0.97) ( 53 ). The combined detection rate was 66%, but when PSA is in the ranges of <0.5, 0.5-0.99, 1.0-1.99, and ≥2 ng/ml, the detection rates were 35%, 41%, 62%, 80%, respectively. Therefore, although the choline tracer is suitable for the detection of BCR of PCa, the detection rate is not ideal when the PSA value is very low. 11 C and 18 F- choline also have implications in assessing prognosis in PCa. Jimbo et al. showed that 11 C-choline PET/CT assessment in mCRPC patients receiving primary docetaxel chemotherapy could predict overall treatment response and progression-free survival with blood pool-corrected SUVmax during treatment ( Figure 2 ) ( 54 ). The percent change in SUVmax was a significant predictor of complete response, with a greater than 20% reduction in SUVmax in 57 of 77 patients (74%), who were 3.6 times more likely to have complete remission than those patients with a reduction of SUVmax <20% after 6 cycles of primary docetaxel chemotherapy. Zhang et al. used 11 C-choline-PET to identify 89 patients with oligometastatic CRPC, providing a better target for stereotactic ablative radiotherapy (SABR) to improve the outcome with a median overall survival of 29.3 months ( 55 ). García Vicente et al. conducted interim and end-of-treatment 18 F-Fluorocholine (FCH) PET/CT imaging in 223 Ra-treated CRPC and bone metastases patients, and the results were significantly associated with both progression-free survival and overall survival, suggesting that interim and end-of-treatment 18 F-FCH PET/CT imaging could be applied as predictors and even guidance during the 223 Ra therapy ( 56 ). 11 C‐choline PET/CT imaging during the docetaxel chemotherapy for a good responder. Baseline (A) , mid‐course (B) , and posttherapy (C) axial fused 11 C‐choline PET/CT images demonstrating markedly choline‐avid right posterior iliac bone metastasis at baseline (arrow), while nearly none at mid-course and posttherapy. Reprinted with permission from Jimbo et al. ( 54 ). Copyright © 2021 Wiley Periodicals LLC. 18 F-fluciclovine ( 18 F-FACBC) was first reported by Shoup in 1999 for brain tumor imaging ( 57 ). Based on the encouraging diagnostic presentation and histologically confirmed data in patients with biochemical recurrence PCa, the FDA and European Commission (EC) approved 18 F-FACBC for diagnostic imaging in PCa patients with elevated PSA after pre-treatment ( 58 ), and until recently, 18 F-FACBC imaging has been included in the National Comprehensive Cancer Network (NCCN) guidelines for the management of BCR of PCa. A previous phase II clinical trial found the sensitivity and specificity of the scan to be 92.5% and 90.1%, respectively, for primary PCa lesions ( 59 ). Uptake of 18 F-FACBC was significantly increased in PCa primary lesions, and lesions with high GS (>3+4) tended to show higher uptake rates compared with low GS lesions and benign prostatic hyperplasia ( 60 ). In the diagnosis of lymph node metastases, this study found that only 1 in 7 patients with metastatic lymph nodes showed true positive results on 18 F-FACBC PET/CT and PET/MRI. Another multicenter phase II study of 40 regional lymph nodes in 28 patients found that the sensitivity, specificity, diagnostic accuracy, positive predictive value, and negative predictive value of 18 F-FACBC imaging in lymph node analysis were 57.1% (4/7), 84.8% (28/33), 80.0% (32/40), 44.4% (4/9) and 90.3% (28/31), respectively ( 61 ). 18 F-FACBC PET/CT imaging has no advantage in the diagnosis of bone metastases either, possibly due to the low spatial resolution and partial volume effects caused by necrotic and mucinous components in the metastatic foci ( 62 ). A meta-analysis included 9 studies and found that the pooled sensitivity and specificity of 18 F-FACBC imaging of aged PCa patients (including both primary and recurrent PCa) were 86.3% and 75.9%, respectively, with a combined diagnostic odds ratio of 16.453 and heterogeneity of 30% ( 63 ). In the regional analysis, 18 F-FACBC-PET/CT owned a higher sensitivity and a lower specificity for the assessment of tumors in the prostate bed than in the extraprostatic region (90.4% vs 76.5%, 89% vs 45%, respectively). Filippi et al. studied the clinical data of 81 patients who underwent 18 F-FACBC PET/CT for BCR of PCa ( 64 ). The detection rate of 18 F-FACBC PET/CT in the entire cohort accounted for 76.9%, and the positive predictive value was 96.7%. This modality played an impact on the clinical management in 33 of 81 patients (40.7%), resulting in a critical amendment in treatment strategy in 30 subjects (90.9%). Like PSMA imaging, the detection rate of FACBC imaging is positively correlated with the PSA levels. When the PSA levels are in the range of 0.2-0.57, 0.58-0.99, 1-1.5 and >1.5 ng/ml, the detection rates of 18 F-FACBC PET/CT were 66.7%, 71.4%, 78.9% and 90, respectively. However, even at a low PSA level, 18 F-FACBC PET/CT imaging preserves a much higher detection rate than PSMA imaging, which is meaningful for the localization and diagnosis of lesions and has a significant impact on clinical management.

Accurate and sensitive imaging using molecular probes is a promising and impactful method for early diagnosis of PCa. In addition, with molecular imaging-based intraoperative guidance, surgeons can achieve precise resection of the malignant PCa tumor as well as the metastatic lymph node, which is the trend in precision medicine. During prostatectomy, including robot-assisted radical prostatectomy (RARP), to maintain the function of the urinary system and erection postoperatively, fluorescent dye or labeled peptide hold great value in enabling visualization and protecting nerve bundles. Although each has disadvantages and limitations, all the novel methods discussed above are essential for developing early diagnosis and effective therapy of PCa. With endless exploration and research, more tracers with higher efficiency will appear to improve the precision theranostic of PCa.

YT, KL and HZ conceived the theme. YT, ZF and YXT conducted the writing of the manuscript. YT and ZF prepared the figures and tables. KL and HZ edited and finalized this manuscript. All authors contributed to the article and approved the submitted version.