The Value of 18F-PSMA-1007 PET/CT in Identifying High-Risk Prostate Cancer

Background: Clinical management decisions on prostate cancer (PCa) are often based on a determination of risk. 68 Ga-prostate-specic membrane antigen (PSMA)-11-positron-emission-tomography (PET)/ computer-tomography (CT) is an attractive modality to assess biochemical recurrence of PCa, detect metastatic disease and stage of primary PCa, making it a promising strategy for risk stratication. However, due to some limitation of 68 Ga-PSMA-11 the development of alternative tracers is of high interest. In this study, we aimed to investigate the value of the new PET trace 18 F-PSMA-1007 in identifying high-risk PCa. Methods: 170 patients with primary PCa underwent 18 F-PSMA-1007 PET/CT were retrospectively analyzed. According to the European Association of Urology (EAU) guidelines on prostate cancer for PCa, patients were classied into low-intermediate-risk (LMR) group or high-risk (HR) group. The maximum standardized uptake values (SUV max ) of the primary prostate tumor was measured on PET/CT images. The diagnostic performance of PET/CT for LMR and HR PCa were calculated and the relationship between the SUV max of primary prostate tumor, prostate-specic antigen (PSA) level and Gleason score (GS) were analyzed. Results: patients into 115 patients were into HR There was a positive correlation between max = 0.597, r = respectively). Tumors and showed lower 18 F-PSMA-1007 GS of and in patients HR LMR (median SUV max : 20.20 versus 8.40; receiver operating characteristic high-risk PCa


Background
PCa is one of the most common tumors in men worldwide [1]. Patients with those high-risk features (de ned by the EAU guidelines on prostate cancer as T2c disease and/or sum Gleason score > 7 and/or serum PSA > 20 ng/ ml) predict a higher risk of metastasis, recurrence or death. The conventional method of identifying high-risk disease in the preliminary diagnosis fails to meet clinical needs. There is a need to develop new methods to allow for appropriate risk strati cation for management, such as active surveillance programs, de nitive therapy, prostatectomy, radiotherapy, or up-front androgen-deprivation.
Incorporation of imaging to current primary PCa classi cations for risk strati cation can help achieve that unmet clinical need [2].
PSMA is a membrane-bound enzyme with high expression in prostate cancer cells and low expression in benign prostatic tissue [3]. Over the past few years, targeted imaging of PSMA has been used in various clinical managements, such as imaging-guided biopsy, staging of primary tumor, localization of biochemical relapse, planning of radiotherapy, prediction, and assessment of tumor response to systemic therapy [4][5][6][7]. The PSMA expression level of PCa and tumor level, Gleason score and PSA stage before treatment had been proved de nitely correlated, and the expression levels have been found to be a predictor for PCa progression [8][9][10]. PSMA-based PET/CT has also been reported to be enabling better tumor detection rate than standard radiologic imaging procedures [11].
Currently, 68 Ga-PSMA-11 is a widely used tracer for PET imaging applications in the detection of PCa.
Nevertheless, the disadvantage of 68 Ga-PSMA PET/CT is that it has more bladder activity, as tracer accumulation in the urinary tract may in uence the uptake evaluation of the prostate bed [12]. Recently, the new PSMA tracer, 18 F-PSMA-1007, can eliminate this kind of disadvantage because of its hepatobiliary excretion owing to its moderate lipophilic characteristics. It has been used as a promising new PET tracer in the management of PCa [13,14]. Furthermore, 18 F-PSMA-1007 has longer half-life and higher physical spatial resolution than 68 Ga-PSMA PET/CT, because 18 F is cyclotron-produced with the larger activity amount [13]. In previous studies, 18 F-PSMA-1007 had been reported that the intensity of tracer accumulation in the primary tumors of PCa patients correlated to GS and PSA level, and it is promising for accurate local staging of PCa [13,15,16]. Furthermore, it has similar or better diagnostic performance than 68 Ga-PSMA-11 in local recurrence or metastasis [14,17]. However, the major limitation of the studies was the relatively small number of patients, and there is limited published data on the diagnosis e cacy of 18 F-PSMA-1007 PET/CT for high-risk PCa.
Thus, we intended to measure the intensity of tracer uptake in the primary prostate tumor and evaluate the value of 18 F-PSMA-1007 PET/CT noninvasive imaging diagnostic strategies to identify the high-risk of PCa and tried to establish an objective imaging reference index.

Patients
In this retrospective study, we analyzed the medical records of 170 PCa patients with Gleason Score of 6 or greater who underwent 18 F-PSMA-1007 PET/CT imaging at our institution between March 2019 and May 2020. The study group included 116 patients with 12-core random, transrectal ultrasound-guided (TRUS) biopsy-proven prostate cancer and 54 patients with radical prostatectomy (RP). In all patients, the time interval between the measurements of PSA values to 18 F-PSMA-1007 PET/CT scan was less than four weeks. Patients were excluded if they received local or systemic treatment, and had the previous history of other cancer. According to the EAU guidelines on prostate cancer [18], all patients were divided into low-intermediate-risk (LMR) group or high-risk (HR) group. The patients of the LMR need to meet all of the following criteria: (1) PSA ≤ 20 ng/ml; (2) Gleason score 6 -7; (3) cTl -cT2b. As the same, the patients of the HR need to meet at least one of the following criteria: (1) PSA > 20 ng/ml; (2) Gleason score 8 -10; (3) above cT2c. Due to the retrospective nature of the study, no formal approval from the ethics committee was required according to our national legislation. Radiopharmaceutical 18 F-PSMA-1007 precursor, cassettes and reagents for the synthesis of 18 F-PSMA-1007 were obtained from ABX advanced biochemical compounds (Radeberg, Germany). 18 F-PSMA-1007 was prepared in a GE TracerLab FN synthesizer according to the one-step procedure described previously [19]. The radiochemical purity of the nal product was >90% as determined by high-performance liquid chromatography. Image analysis All 18 F-PSMA PET/CT images were analyzed using a dedicated workstation (EBW3.0, Philips), which allowed the review of PET, CT and fused imaging data in axial, coronal and sagittal slices. PET imaging was interpreted independently by 2 experienced nuclear medicine physicians both of whom have more than 10 years of clinical experience and blind of all relevant clinical statistics. Any disagreement was resolved by consensus. SUV max of the primary tumors were acquired from the most intense uptake area in prostate gland. Areas in the whole body having uptake above the background activity were de ned as metastatic. Typical pitfalls such as PSMA uptake in sacral and coeliac ganglia or in the stellate ganglia were frequently observed but were not considered pathological [20]. This interpretation criterion comes from the result of our clinical experience and consistent with published literature [21][22][23][24].

Statistical analysis
Data analyses were performed with SPSS version 23.0 software (SPSS, Chicago, IL). Associations between GS, PSA value, and SUV max of the primary tumor were described descriptively (Nonparametric Spearman correlation coe cients). The differences between different subgroups were evaluated by using the Mann-Whitney U test and Kruskal-Wallis test. ROC curve analysis was used to determine the optimal cutoff value of the SUV max for identifying high-risk PCa. For all statistical parameters, P values of less than 0.05 were considered statistically signi cant.

Patients' Characteristics
The clinical characteristics of the enrolled 170 patients with GS 6 to 10 are summarized in Table 1  There was a statistically signi cant difference in median SUV max between patients of HR and those of LMR (20.20 vs. 8.40, P < 0.001; Table 2). For the Gleason score, the detailed information about the SUV max values of different GS subgroups are summarized in Table 3. Gleason score and SUV max of primary tumors showed a signi cant positive correlation with each other (r = 0.446, P < 0.001).
Combining GS and tumor-related tracer uptake, lower median SUV max value was found in the subgroups

Discussion
In this retrospective study, we found that there is a certain positive correlation between the intensity of 18 F-PSMA-1007 accumulation and the GS/PSA level in the primary tumors of PCa patients. Furthermore, the SUV max of the primary tumor was valuable for identifying high-risk PCa.
A timely and accurate diagnosis of high-risk PC is front and center for the clinician. The commonly used risk classi cation for the PC are based on clinical stage, Gleason score by biopsy, and PSA level before treatment. However, it is not absolutely reliable to evaluate the accuracy of GS in patients who have undergone TURS. In the clinical work, it may also encounter the patient who refuse biopsy for a variety of reasons. Another problem with the scheme is the inherent inaccuracy in determining T stage [2]. Assessing disease by digital rectal examination has signi cant inter-observer variability. PSMA-PET/CT, as a noninvasive imaging diagnostic strategy, may compensate for these shortcomings. Recent studies found a statistically signi cant positive correlation between GS/PSA value and SUV max of primary tumors on PSMA-PET/CT [13,23]. Kesch et al. proved 18 F-PSMA-1007 PET/CT and multiparametric magnetic resonance imaging had similar diagnostic performance in local staging of PCa [15]. In our study, the SUV max showed a signi cant association with the presence of high-risk PCa. Patients of HR had signi cantly higher SUV max than those of LMR (P < 0.001). The AUC of the SUV max of the primary tumor is 0.873, which can e caciously identify high-risk patients with PCa. Therefore, we believe pathologists and clinicians may reduce missed diagnoses if they refer to PET images and results. Besides, PSMA-PET/CT may better screen out the patients of high risk, especially when the patients are unable to receive aspiration biopsy or the histology results of biopsy are not satisfactory.
The biological characteristics of PCa tissues vary greatly between different GS, which is an important indicator for the treatment and prognosis evaluation of PCa [18]. Thus, we also made the pairwise comparison between different GS subgroups and found that there were statistically signi cant differences in SUV max between the subgroups of GS 6 and GS 7a, and the subgroups of GS 8-10 (P < 0.001). There was no statistical difference in SUVmax value between tumors with 7b and those with GS of 8-10, which was different from the result of previous studies on 68 Ga-PSMA [23,24]. Reasons for these discrepancies remain speculative. The pathological results of TRUS-biopsy may underestimate the actual Gleason scores, which might be one of the reasons. Previous study had shown that compared with the subgroup GS 7b, the dangerous level of the subgroup GS 7a tumors could be treated conservatively without the need for a radical surgery [25], thus, the distinguishment between the two subgroups is of great importance for clinical treatment. But it is worth noting that the SUV max of primary tumor between these two subgroups has no statistical difference with a median SUV max of 8.76 (GS 7a) and 14.30(GS 7b), (P > 0.05). This nding was consistent with previous studies on 68 Ga-PSMA [23,24], which may Figure 4 showed the result of the ROC curve analysis for high-risk PCa. The AUC of the SUV max was 0.873. The sensitivity and speci city were 90.4% and 69.1%, respectively. The optimal cut-off values of SUV max was set as 10.78.
reveal that the stage difference between GS 7a and GS 7b was not enough to cause a difference in SUV max on PSMA-PET/CT.
The present study had some limitations that should not be neglected. Firstly, the retrospective nature of the analysis is the major limitation of our study, and further validation is required by multicenter studies with more patients. Secondly, the patients of histopathological con rmation account for a larger proportion of all patients, thus, there might be a possibility that the histology results of primary tumors were underestimated in some of the enrolled patients. Finally, histopathological con rmation was not acquired for metastatic lesions in this study.

Conclusion
In conclusion, 18  All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Due to the retrospective nature of the study, no formal approval from the ethics committee was required according to our national legislation.

Consent for publication
Not applicable Availability of data and material The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.   Comparison of 18F-PSMA-1007 uptake expressed in SUVmax value in primary tumors of different GS subgroups. Box plots demonstrate that higher GS exhibited statistically signi cant higher tracer uptake in the primary tumor.