18F-PSMA-1007 PET in Biochemical Recurrent Prostate Cancer: An Updated Meta-Analysis

Background Prostate-specific membrane antigen- (PSMA-) targeted agents labeled with fluorine-18 (18F) have recently become available to evaluate patients with biochemical recurrent prostate cancer (BRPCa) by using positron emission tomography/computed tomography (PET/CT) or positron emission tomography/magnetic resonance imaging (PET/MRI). We performed a systematic review and meta-analysis about the detection rate (DR) of 18F-PSMA-1007 PET/CT or PET/MRI in BRPCa patients. Methods A comprehensive computer literature search of PubMed/MEDLINE, EMBASE, and Cochrane Library databases for studies published through 17 May 2021 was carried out using the following search algorithm: “PSMA” AND “1007”. Only studies providing data on the DR of 18F-PSMA-1007 PET/CT or PET/MRI in BRPCa were included. A random-effects model was used to calculate the pooled DR on a per scan basis. Results Fifteen articles (853 patients) were selected and included in the systematic review, and ten were included in the quantitative analysis. Most of the studies reported a good DR of 18F-PSMA-1007 PET/CT or PET/MRI in BRPCa including also patients with low prostate-specific membrane antigen (PSA) values. The DR of 18F-PSMA-1007 PET/CT or PET/MRI was dependent on PSA serum values. The pooled DR was 81.3% (95% confidence interval: 74.6–88%) with statistical heterogeneity. A significant reporting bias (publication bias) was not detected. Conclusions 18F-PSMA-1007 PET/CT or PET/MRI showed a good DR in BRPCa patients in line with other PSMA-targeted agents. The DR of 18F-PSMA-1007 PET/CT or PET/MRI is influenced by serum PSA values. These findings should be confirmed by prospective multicentric trials.


Introduction
Between 27% and 53% of all patients with prostate cancer (PCa) undergoing radical prostatectomy (RP) or radiation therapy (RT) develop a biochemical recurrence (BCR) defined as rising prostate-specific antigen (PSA) serum levels after curative treatment for PCa [1]. Metabolic imaging methods and in particular hybrid positron emission tomography/computed tomography (PET/CT) or positron emission tomography/magnetic resonance imaging (MRI) using prostate-specific membrane antigen-(PSMA-) targeted agents have been recognized as useful techniques in improving the diagnosis of biochemical recurrent prostate cancer (BRPCa) [2][3][4][5]. e identification of sites of BRPCa can affect treatment decisions as therapies in PCa patients with local recurrence or distant metastatic spread are different [1].
PSMA is a protein overexpressed in PCa tumor cells, and this is the rationale for using PSMA-targeted agents in the diagnosis and therapy of PCa [6,7]. However, PSMA is even overexpressed in other tumors beyond PCa and in other organs and tissues [6,7].
After successful preclinical studies [14], 18 F-PSMA-1007 has been introduced in clinical practice. is biodistribution of this PET radiopharmaceutical is similar compared to other PSMA-targeted agents [13][14][15][16]. e increased uptake of this radiopharmaceutical in PCa lesions compared with other PSMA-targeted tracers improves tumor-to-background ratios, facilitating the detection of small PCa lesions. Another clear advantage of 18 F-PSMA-1007 compared to other PSMA-targeted radiopharmaceuticals is its predominant excretion via the hepatobiliary rather than urinary system, allowing an excellent assessment of the pelvic region owing to the reduced interference of the urinary radioactivity [13][14][15][16].
A recent systematic review demonstrated the good detection rate of 18 F-PSMA-1007 PET/CT in staging PCa [15]. Our aim is now to perform an updated meta-analysis to calculate the detection rate of 18 F-PSMA-1007 in BRPCa.

Methods
e Preferred Reporting Items for a Systematic Review and Meta-Analysis of Diagnostic Test Accuracy Studies" (PRISMA-DTA) statement and other specific guidelines [17,18] were used to write this evidence-based article.

Creation of a Review Question and Related Search Strategy.
First, the following review question was created: "Which is the DR of 18 F-PSMA-1007 PET/CT or PET/MRI in patients with BRPCa?". Based on the review question, the authors independently performed a literature search using several bibliographic databases (PubMed/MEDLINE, EMBASE, and Cochrane library) to find articles related to the review question. e search string was "PSMA" AND "1007". No filters about date and language were used. e last update of the literature search was 17 May 2021. e authors also screened the references of the retrieved articles to find further eligible studies.

Selection of Studies.
Studies or subsets of studies investigating the DR of 18 F-PSMA-1007 PET/CT or PET/MRI in BRPCa patients were eligible for inclusion in the qualitative (systematic review) and quantitative analysis (metaanalysis). e exclusion criteria for the systematic review were as follows: (a) articles not within the field of interest of this review (including articles that evaluated 18 F-PSMA-1007 PET/CT or PET/MRI for PCa staging or those who evaluated BRPCa patients with other PET radiopharmaceuticals beyond 18 F-PSMA-1007); (b) review articles, editorials, comments, letters, and conference proceedings related to the review question; and (c) case reports or small case series related to the review question (less than 5 BRPCa patients included). For the quantitative analysis, the articles with data overlap or without sufficient data to calculate the DR on a per scan-based analysis were excluded from the meta-analysis but included in the qualitative analysis (systematic review).
Two researchers (MF and GT) independently reviewed abstracts, titles, and full-text of the retrieved studies. e abovementioned inclusion and exclusion criteria were applied, and disagreements were discussed and solved in an online meeting.

Extraction of Data.
Two researchers (MF and GT) independently extracted the data from the selected articles. For each article, several pieces of information were collected in particular about basic study characteristics (name of authors, publication year, country of origin, and study design), characteristics of patients included (e.g., number and type of BRPCa patients scanned with 18 F-PSMA-1007 PET/CT or PET/MRI, mean/median age, PSA serum values and PSA kinetics, Gleason score). Furthermore, data on technical aspects were extracted (e.g., hybrid imaging technique and PET scanner, mean injected activity, the time between 18 F-PSMA-1007 injection and PET acquisition, PET image analysis, verification of PET/CT and PET/MRI findings, and other imaging modalities performed for comparison). For articles included in the analysis, information was collected about DR values of 18 F-PSMA-1007 PET/CT or PET/MRI (overall estimates and at different PSA cut-off values) on a per scan-based analysis, mean PSA serum values in patients with positive and negative 18 F-PSMA PET/CT, and percentage of change of management by using 18 F-PSMA-1007 PET/CT or PET/MRI in BRPCa patients.

Quality Assessment.
e overall quality of the studies included in the systematic review was critically appraised based on the revised "Quality Assessment of Diagnostic Accuracy Studies" tool (QUADAS-2) [19].

Statistics.
e DR was defined as the ratio between the number of scans with at least one suspected lesion detected and the total number of scans performed. Pooled analysis about DR of 18 F-PSMA-1007 PET/CT or PET/MRI was performed using data retrieved from the included studies. A random-effects model (as suggested by DerSimonian and Laird) was used for the pooled analysis [18]. Pooled estimates and related 95% confidence interval (95%CI) values were calculated. Forest plots were provided for the metaanalysis. e I-square index (I 2 ) or inconsistency index was used to estimate the statistical heterogeneity [18,20], whereas the publication bias was assessed through the visual analysis of funnel plot and using Egger's test [18,21]. e open-source OpenMetaAnalyst software developed by the Center for Evidence Synthesis in Health of Brown University (USA) was used for the statistical analysis.

Literature Search.
Literature search results are reported in Figure 1. e comprehensive computer literature search from PubMed/MEDLINE, EMBASE, and Cochrane library database revealed 98 records. Reviewing titles and abstracts, 83 records were excluded: 70 because not in the field of interest of this review, 7 as reviews, editorials, or letters, and 6 as case reports. Fifteen articles were selected, and their fulltext version was retrieved [22][23][24][25][26][27][28][29][30][31][32][33][34][35][36]. No additional studies were found screening the references of these articles. erefore, 15 articles were eligible for the qualitative analysis (systematic review). ree articles were excluded from the meta-analysis for possible patient data overlap [24,28,34] and two for insufficient data to calculate the DR on a per scan-based analysis [32,36]. Finally, 10 articles were included in the quantitative analysis (meta-analysis) [22, 23, 25-27, 29-31, 33, 35]. Tables 1-3 show the characteristics of the selected studies and their main findings. Figure 2 shows the overall quality assessment of the selected studies.

Characteristics of Studies and Patients Included.
rough the literature search using the selected bibliographic databases, 15 full-text articles (853 BRPCa patients) were selected [22][23][24][25][26][27][28][29][30][31][32][33][34][35][36]. All the eligible articles were recently published; different countries were represented; most of the studies were retrospective (73%) or monocentric (87%), and there were no prospective multicentre studies. e mean/ median age of BRPCa patients included ranged from 67 to 71 years. e Gleason score distribution largely varied across the studies. e mean and median PSA serum values of BRPCa patients before PET/CTor PET/MRI ranged from 0.6 to 11 ng/mL. Unfortunately, only a limited number of selected articles reported information about PSA kinetics. Table 2. e hybrid imaging modality was PET/CT in 14/15 studies (93%) and PET/MRI in one study (7%). e mean injected radiotracer activity and the time interval between radiotracer injection and image acquisition were quite different among the included studies. e PET image analysis was performed by using visual analysis in all studies; additional calculation of standardized uptake values (SUV) was performed in about half of the studies (8/15; 47%). Quantitative analysis was also performed by one study. At visual analysis, lesions with a focal 18 F-PSMA-1007 uptake higher than the local background activity, excluding physiological uptake areas or known pitfalls of 18 F-PSMA-1007 PET, were considered positive for PCa. In one study, PET reviewers used the criteria for harmonization of PSMA PET/CT interpretation [23].
e results about the correlation between the Gleason score and DR of 18 F-PSMA-1007 PET in BRPCa patients were controversial. In one study, in PCa patients with a higher Gleason score of the primary tumor (≥8), the DR of 18 F-PSMA-1007 PET trended higher compared to those with lower Gleason score (<8) but without a statistically significant difference [25]. In one study, the Gleason score was found as a significant independent predictor of bone metastases on 18 F-PSMA-1007 PET/CT [22]. In two studies, DR of 18 F-PSMA-1007 PET/CT was independent of the Gleason score [29,35].
When compared to CT, a significantly higher number of positive findings for PCa were detected by using 18 F-PSMA-1007 PET/CT [27].
When compared to PET/CT with renally excreted PSMA ligands, the interpretation of locoregional PSMA-positive lesions was increased by using 18 F-PSMA-1007 PET imaging; to this regard, the very low urinary activity in 18 F-PSMA-1007 PET scans, due to the minimal clearance via the urinary pathway, seems to be a clear advantage in favor of 18 F-PSMA-1007 [23,28,30,36]. However, a significant number of unspecific findings in the bone marrow may be detected by 18 F-PSMA-1007 PET [23]. Furthermore, compared to 68 Ga-PSMA-11 PET/CT, 18 F-PSMA-1007 PET/CT detected a higher number of benign lesions without a significant difference in DR of BRPCa lesions among these imaging methods [30]. Conversely, large equality between 18 F-DCFPyL and 18 F-PSMA-1007 PET/CT was found even if 18 F-PSMA-1007 PET/CT detected suspected local relapse in a higher proportion of BRPCa patients whereas 18 F-DCFPyL PET/CT showed less equivocal skeletal lesions [36].
DR of 18 F-PSMA-1007 PET/CT in BRPCa patients was significantly higher compared to 18 F-choline PET/CT with a Contrast Media & Molecular Imaging higher number of BRPCa lesions detected and a lower number of negative and equivocal results [34].
In some studies, the change of management by using 18 F-PSMA-1007 PET/CT or PET/MRI in BRPCa was reported and it ranged from 64% to 79% of cases [24,26,33].

Discussion
Our report is the first systematic review and meta-analysis specifically focused on 18 F-PSMA-1007 PET/CT or PET/ MRI in BRPCa patients. In a previous meta-analysis, data from different 18 F-labeled PSMA tracers used in BRPCa patients were pooled together [16]. We believe that, taking also into account the increasing literature data, an updated systematic review and meta-analysis focused on a specific 18 F-labeled PSMA radiopharmaceutical, as performed in this manuscript, should be preferred because each PSMA-related radiopharmaceutical has different characteristics which can lead to different DR.
Compared to other imaging methods, 18 F-PSMA-1007 PET/CT or PET/MRI clearly showed a higher DR than conventional cross-sectional imaging with CT. is finding is not surprising, as functional abnormalities detected by functional imaging techniques usually precede morphological abnormalities detected by morphological imaging. Overall, CT alone seems insufficient for restaging of BRPCa patients [27]. e pooled DR of 18 F-PSMA-1007 PET/CT or PET/MRI in BRPCa is similar compared to 68 Ga-labeled PSMA PET/ CT or PET/MRI [3]. However, the longer half-life and higher injected activities of 18 F-labeled PSMA tracers should allow higher lesion uptake, superior clearance of background activity, and higher quality images of PET with 18 F-labeled PSMA tracers compared to PET with 68 Ga-labeled PSMA tracers [15,16]. Furthermore, the lower positron energy of 18 F than 68 Ga theoretically provides an improved spatial resolution [38].    Only two studies compared 18 F-PSMA-1007 and 68 Ga-PSMA-11 PET/CT in BRPCa patients [23,30]. A similar number of lesions was attributed to recurrent PCa using 18 F-PSMA-1007 and 68 Ga-PSMA-11, whereas a significantly higher number of benign lesions were detected using 18 F-PSMA-1007 [30]. Assuming similar DR, the real advantage of 18 F-PSMA-1007 compared to 68 Ga-PSMA-11 could be the large-scale production by a cyclotron and supply to PET centers due to the longer half-life of 18 F (110 min) compared to 68 Ga (68 min) [15,16].
Compared to other 18 F-labeled PSMA tracers, 18 F-PSMA-1007 may provide a better interpretation of lesions adjacent to the urinary tract but may decrease the interpretability of skeletal lesions due to the higher frequency of focal areas of unspecific radiotracer uptake in the bone [23,36]. ese findings underline that the main advantage of 18 F-PSMA-1007 compared to other PSMA-targeted PET tracers is its reduced urinary clearance allowing an excellent assessment of the pelvic region owing to the reduced interference of the urinary radioactivity [15,16]. On the other hand, focal unspecific bone uptake on 18 F-PSMA-1007 PET seems a frequent finding (more frequent on digital PET scanners than analog PET scanners), and it should be interpreted carefully to avoid PCa overstaging [39,40].
Compared to 18 F-choline, DR of 18 F-PSMA-1007 PET/ CT in BRPCa patients was significantly higher with a higher number of BRPCa lesions detected and a lower number of negative and equivocal results [34]. is is in line with the findings of a previous meta-analysis which demonstrated that PSMA-targeted PET/CT was clearly superior in detecting BRPCa lesions when compared to radiolabeled choline PET/CT, in particular in the subgroup of BRPCa patients with low PSA levels (≤1 ng/ml) [41].
Only three articles evaluated the change of management using 18 F-PSMA-1007 PET/CT or PET/MRI in BRPCa [24,26,33]. Interestingly, a significant percentage of change of management was reported in these studies ranging from 64 to 79% of cases. ese findings are in line with those reported using PET with other PSMA-targeted agents in BRPCa [42]. e main limitations of our systematic review and metaanalysis are the limited number of studies included, the verification bias in the included studies (not all positive PET lesions were verified by histology), and the heterogeneity among studies. Furthermore, the different interpretation criteria could have an influence on the DR of 18 F-PSMA-1007 PET/CT or PET/MRI. In this regard, a standardized interpretation of PSMA-ligand PET has been recently proposed [43].
We have detected statistical heterogeneity among the included studies in our meta-analysis, and unfortunately, there were insufficient data to perform meta-regression and several subgroup analyses. Only a small subgroup analysis taking into account a PSA cut-off value of 0.5 ng/mL was performed. On the other hand, we did not find a significant publication bias. e hybrid imaging modality used in the included studies was PET/CT in most of the articles and PET/MRI only in one study; therefore, we need more data to further evaluate the DR of 18 F-PSMA-1007 PET/MRI in comparison to 18 F-PSMA-1007 PET/CT. Lastly, large prospective multicentre studies and, in particular, cost-effectiveness analyses comparing 18 F-PSMA-1007 to other PET radiopharmaceuticals are warranted to confirm these findings. Data Availability e underlying data supporting the results of our systematic review article can be found searching the scientific literature database (PubMed, Embase, and Cochrane Library).

Conflicts of Interest
e authors declare no conflicts of interest.

Authors' Contributions
Conceptualization was done by GT and MF; methodology was developed by GT; software was provided by GT; data curation was performed by MF and GT; original draft was written by GT; reviewing and editing were performed by MF.