Preliminary Results of an Ongoing Prospective Clinical Trial on the Use of 68Ga-PSMA and 68Ga-DOTA-RM2 PET/MRI in Staging of High-Risk Prostate Cancer Patients

The aim of the present study is to investigate the synergic role of 68Ga-PSMA PET/MRI and 68Ga-DOTA-RM2 PET/MRI in prostate cancer (PCa) staging. We present pilot data on twenty-two patients with biopsy-proven PCa that underwent 68Ga-PSMA PET/MRI for staging purposes, with 19/22 also undergoing 68Gaa-DOTA-RM2 PET/MRI. TNM classification based on image findings was performed and quantitative imaging parameters were collected for each scan. Furthermore, twelve patients underwent radical prostatectomy with the availability of histological data that were used as the gold standard to validate intraprostatic findings. A DICE score between regions of interest manually segmented on the primary tumour on 68Ga-PSMA PET, 68Ga-DOTA-RM2 PET and on T2 MRI was computed. All imaging modalities detected the primary PCa in 18/19 patients, with 68Ga-DOTA-RM2 PET not detecting any lesion in 1/19 patients. In the remaining patients, 68Ga-PSMA and MRI were concordant. Seven patients presented seminal vesicles involvement on MRI, with two of these being also detected by 68Ga-PSMA, and 68Ga-DOTA-RM2 PET being negative. Regarding extraprostatic disease, 68Ga-PSMA PET, 68Ga-DOTA-RM2 PET and MRI resulted positive in seven, four and five patients at lymph-nodal level, respectively, and at a bone level in three, zero and one patients, respectively. These preliminary results suggest the potential complementary role of 68Ga-PSMA PET, 68Ga-DOTA-RM2 PET and MRI in PCa characterization during the staging phase.


Introduction
Prostate cancer (PCa) is one of the worldwide leading causes of cancer-related death. Approximately 15% of men present with high-risk PCa, which is characterized by an increased risk of extracapsular extension, locally advanced disease, and/or bone metastases [1]. Hence, at diagnosis, a whole-body staging for high-risk PCa patients is strongly recommended regardless of the surgical or radiation-based treatment decision [2].
The current staging of intermediate and high-risk PCa includes imaging of abdomen and pelvis performed by using Computed Tomography (CT) or Magnetic Resonance Imaging (MRI) and bone scan to evaluate potential sites of metastatic spread.
The current EAU-ESTRO-SIOG guidelines report that Positron Emission Tomography/CT (PET/CT) is a valuable imaging modality that might be considered in men with high-risk diseases undergoing initial staging [3]. However, as no randomised-control trials demonstrating survival benefit are available yet, its role in guiding therapeutic decisions must be cautious [3].
Multi-parametric MRI (mp-MRI) is a well-established imaging modality for PCa assessment and it is used to detect the primary tumour, guide biopsies and define the local extent of the disease; its usefulness for local staging has been largely reported, although local staging with MRI might be associated with limited sensitivity [4][5][6].
Molecular imaging with PET represents a valid imaging approach in PCa staging, with new PET tracers other than Choline having a relevant role in improving diagnoses, staging and follow-up of PCa [7][8][9][10].
In this regard, prostate-specific membrane antigen (PSMA), a transmembrane protein with a significantly increased expression in PCa cells, is an imaging probe that has been introduced in clinical practice, with recent data demonstrating good accuracy in PCa staging [11,12].
Gastrin releasing peptide receptor (GRPR) is a G-protein coupled receptor overexpressed in different types of cancer including PCa [13,14]. The 68 Ga-DOTA-RM2 is a GRPR antagonist used as a PET imaging probe that has demonstrated promising, but still limited results in PCa imaging [15][16][17].
Hybrid PET/MRI allows for the simultaneous acquisition of metabolic, structural, and functional imaging information regarding PCa status in a whole-body single session examination, thus representing an innovative imaging approach capable to overcome the pitfalls of conventional imaging and, potentially, helping clinicians in the management of PCa. Only a few, preliminary studies have compared 68 Ga-PSMA and 68 Ga-DOTA-RM2-PET radiotracers in PCa by using PET/CT or PET/MRI so far, with promising results in both patients presenting with biochemical recurrence and in those with newly diagnosed intermediate-or high-risk prostate cancer [15,18,19].
The aim of the present study is to report our preliminary experience on the synergic use of 68 Ga-PSMA PET/MRI and 68 Ga-RM2 PET/MRI in prostate cancer staging.

Patients
In this prospective clinical study, 22 patients with biopsy-proven PCa were enrolled from 1 September 2020 to 31 August 2021 at the IRCCS San Raffaele Scientific Institute.
Inclusion criteria were age greater than 18 years at the time of PET/MRI scan, biopsyproven high-risk PCa (defined as PSA > 20 ng/mL and/or clinical stage ≥ cT2c and/or biopsy ISUP grade ≥ 4, according to European Association of Urology guidelines [3]) candidate to prostatectomy and pelvic lymphadenectomy. Exclusion criteria were inability to complete the required imaging examinations (i.e., severe claustrophobia), medical condition possibly interfering and significantly affecting study compliance, all contraindications to undergo MRI scan (i.e., metallic/conductive or electrically/magnetically active implants without MR-safe or MR-conditional labelling) and evidence of metastatic disease on conventional imaging contraindicating the surgical procedure.
All recruited patients underwent 68 Ga-PSMA PET/MRI, with 19 also undergoing 68 Ga-DOTA-RM2 PET/MRI in two different days, with at least 48h interval, for staging purposes before radical prostatectomy. Histological validation of imaging findings was Diagnostics 2021, 11,2068 3 of 20 retrieved from clinical reports for all patients who have undergone radical prostatectomy, so far.
This study was approved by the Institutional Ethics Committee of IRCCS San Raffaele Scientific Institute (EudraCT: 2018-001034-18) and all patients gave written informed consent to participate in the study.

68 Ga-PSMA PET/MRI Acquisition Protocol
68 Ga-PSMA-11 was synthesised by a fully automated synthesis module (Neptis Mosaic-RS, ORA, Neuville, Belgium) connected to a 68 Ge/ 68 Ga generator (1.85 GBq Galli Ad, IRE ELiT, Fleurus, Belgium) and equipped with a disposable single-use cassette kit (ABX GmbH, Radeberg, Germany). A standardised labelling sequence with 15 µg (15 nmol) of unlabelled PSMA 11 (ABX GmbH) was used. The final product was sterilely filtered over 0.22 µm PVDF filters. For quality control, 68 Ga-PSMA 11 was analysed by radio analytic high-performance liquid chromatography on a modular system (Waters) equipped with a diode array detector and a radio detector using an RP-18 column (ACE 5 µm C18, 150 × 3 mm, Advanced Chromatography Technologies Ltd., Aberdeen, Scotland). A gradient elution over 13 min at a flow of 1.5 mL/min from 90%A to 30%A and again 90%A was employed, where Solvent A was Water + 0.1% TFA and Solvent B was CH3CN+ 0.1% TFA. Other quality controls performed before release included TLC on iTLC strips with MeOH/1M AcONH4, pH measurement and radionuclidic purity. Residual HEPES determination, Ethanol quantification and Microbiological purity were assessed on decayed product. Uncorrected radiochemical yield was over 70% with a radiochemical purity > 91%.
The 68G a-PSMA PET/MR examination protocol included a high statistic (HS) scan (20 min), covering a single bed position, that was simultaneously acquired to the following MR sequences: Following the single bed acquisition, a total-body (TB) PET scan (5-6 FOVs, 4 min/FOV) was then simultaneously acquired to an MRI TB T1 Lava Flex sequence and a TB DWI with b = 50, b = 1000 s/mm 2 .
PET images were reconstructed using a Bayesian penalised likelihood reconstruction algorithm [20] with a reconstructed FOV of 60 cm and image matrix of 192 × 192. The algorithm includes a Point Spread Function and Time of Flight information. Attenuation Correction (AC) of PET data was performed using MR AC technique based on the processing of the LAVA-Flex sequences acquired simultaneously with the PET data.

The 68 Ga-DOTA-RM2 PET/MRI Acquisition Protocol
The 68 Ga-DOTA-RM2 was synthesised by a kit-like procedure developed for the radiolabelling with GalliAd ® generator (IRE Elite, Fleurus, Belgium). Briefly, the eluate from the 68 Ge/ 68 Ga generator (1.85 GBq Galli Ad, IRE ELiT, Fleurus, Belgium) was added to a sterile vial containing 40 µg of DOTA-RM2 (Life Molecular Imaging, Fribourg, Germany) in format buffer and ascorbic acid. Reaction vial is placed in a pre-heated thermostat at 115 • C for 10 min. Successively, vial is left to cool down for 10 min at room temperature. No purification step was needed. The solution is sterile filtered over sterile 0.22 µm PVDF membrane and dispensed as injectable solution. For quality control, 68 Ga-DOTA-RM2 was analysed by radio analytic high-performance liquid chromatography on a modular system (Waters) equipped with a diode array detector and a radio detector using an RP-18 column (ACE 5 µm C18, 150 × 3 mm, Advanced Chromatography Technologies Ltd., Aberdeen, Scotland). A gradient elution over 16 min at a flow of 1.0 mL/min from 80%A to 20%A and again 80%A was employed, where Solvent A was Water + 0.1% TFA and Solvent B was CH3CN + 0.1% TFA. Unbound gallium was quantified by iTLC strips with MeOH/1M AcONH4 while ionic gallium with TLC strips with 0.1 M sodium citrate pH5. pH of the final solution was 3.2-3.8. Radionuclidic purity was assessed before release, microbiological purity was assessed on decayed product. Uncorrected radiochemical yield was over 60% with a radio-chemical purity > 91%.
As for 68 Ga-PSMA PET/MRI, fasting condition was requested on the day of the examination and the same PET/MRI scanner was used.
The 68 Ga-DOTA-RM2 PET/MR examination protocol included an HS scan (20 min), covering a single bed position, that was simultaneously acquired to the following MR sequences: an axial T2 weighted sequence with large FOV (32 × 32 cm 2 ), an axial 3D T2 sequence with small FOV, a T1-Lava Flex sequence of the whole pelvic region. Following the single bed acquisition, a TB PET scan (5-6 FOVs, 4 min/FOV) was then simultaneously acquired with a TB axial Lava Flex sequence and a TB sagittal STIR sequence on the spine. Reconstruction and attenuation correction of PET images were performed by using the same algorithms and parameters used for 68 Ga-PSMA PET images.

PET/MR Image Analysis
A 68 GA-PSMA and 68 Ga-DOTA-RM2 image read-out was performed on the Advantage Workstation (AW, General Electric Healthcare, Waukesha, WI, USA) on which PET, MRI and fused PET/MRI images could be visualized in axial, coronal and sagittal planes. HS PET acquisition bed on the pelvic region and TB PET examination of both 68 Ga-PSMA and 68 Ga-DOTA-RM2 PET images were qualitatively interpreted by two experienced (more than 10 years of experience) Nuclear Medicine physicians, with knowledge of all the available patients' clinical and imaging information.
For primary tumour assessment, HS and TB pelvic PET images were qualitatively evaluated for both 68 Ga-PSMA and 68 Ga-DOTA-RM2. The presence of 68 Ga-PSMA and 68G a-DOTA-RM2 increased uptake was considered positive for malignancy with the anatomical site being defined on the basis of MRI anatomy, except for those areas of physiologically increased uptake [21,22]. Regions of interest (ROIs) on the primary tumour, showing 68 Ga-PSMA and 68 Ga-DOTA-RM2 uptake on HS PET images were semi-automatically defined on transaxial PET images. Furthermore, the following semi-quantitative parameters were calculated for the primary tumour on HS PET images for both radiotracers: maximum standardised uptake value (SUVmax), mean SUV (SUVmean, using different thresholds, namely 40%, 50%, 60% of the maximum value-SUVmean40, SUVmean50, SU-Vmean60) and metabolic tumour volume (MTV) calculated at different thresholds (MTV40, MTV50, MTV60).
In addition, to determine the volume and the location of 68 Ga-PSMA and 68 Ga-DOTA-RM2 PET primary tumour uptake, one experienced Nuclear Medicine physician manually segmented the primary tumour slice-by-slice using 3D Slicer software (revision 29402) [23] on both 68 Ga-PSMA and 68 Ga-DOTA-RM2 PET images. Afterward, all segmentations were co-registered and brought to a common reference volume (one of the first PET studies). To do that, the MRAC of the 68 Ga-DOTA-RM2 PET study was firstly co-registered to the one of the 68 Ga-PSMA PET, by means of 3D Slicer, and then the obtained transformations were applied to the 68 Ga-DOTA-RM2 PET images and the corresponding ROI segmentation.
After the evaluation of the primary prostatic tumour, the whole-body distribution pattern of both 68 Ga-PSMA and 68 Ga-DOTA-RM2 were qualitatively assessed, and the presence of extra-prostatic 68 Ga-PSMA and 68 Ga-DOTA-RM2 increased uptake was considered positive for malignancy, with the exception of areas of physiologically increased uptake. The number and the site of lymph nodal involvement were reported, as well as the presence of suspect distant metastases. The anatomical site was defined on the basis of MRI anatomy.
In case of suspect bone metastasis in PET images, the whole-body MRI sequences were screened to confirm the spreading of the disease.
MR images acquired during HS 68 Ga-PSMA PET were initially processed using AW software: small FOV DWI with b values of 50-800 were used to generate ADC maps. Volumetric ROIs of lesions visible to T2 and ADC images were created using 3D Slicer to obtain the following quantitative parameters: lesion volume, mean ADC (ADCmean) and minimum (ADCmin).

Qualitative and Quantitative Comparison of 68 Ga-PSMA, 68 Ga-DOTA-RM2 and MRI
A qualitative comparison between 68 Ga-PSMA and 68 Ga-DOTA-RM2 intra-prostatic uptake and morphological findings detected on MR images was performed in order to describe the possible concordances and discrepancies between metabolic and morphologic imaging. Moreover, a qualitative comparison was also performed in terms of number and sites of lymph nodal and distant metastases for all patients, considering all three different imaging modalities.
Finally, DICE score between the ROIs manually segmented on the primary tumour on 68 Ga-PSMA PET and 68 Ga-DOTA-RM2 PET and on MRI was computed in order to evaluate the correspondence of the intra-prostatic findings referable to the site of primary tumour across modalities.

Correlations between PET Semi-Quantitative and MRI Quantitative Imaging Parameters
To provide improved characterization of the primary tumour, a correlation between multitracer PET and MRI parameters was performed. In particular, a Spearman correlation was calculated between the semi-quantitative PET parameters measured on HS 68 Ga-PSMA PET and HS 68 Ga-DOTA-RM2 PET images (SUVmax, SUVmean40, SUVmean50, SUVmean60, MTV40, MTV50, MTV60, manually segmented lesion volume), the quantitative parameters measured on MRI (manually segmented lesion volume, ADCmin, ADCmean) and clinical data (PSA, Gleason Score, ISUP score). A p-value < 0.05 was considered statistically significant.

Patients
Twenty-two men (mean age: 65 years; range 52-80) with biopsy-proven high-risk PCa were enrolled in this prospective pilot study, so far. All patients had a Gleason score ≥ 7 on biopsy, with a mean PSA at time of diagnosis of 7.40 ng/mL (range: 2.5-26.17). Patients' characteristics are reported in Table 1. All patients underwent 68 Ga-PSMA PET/MRI and 19/22 also underwent 68 Ga-DOTA-RM2 PET/MRI within sixteen days (mean: 3 days, range: 2-16 days) because of reduced compliance to the study protocol. Twelve patients have undergone radical prostatectomy so far, with the availability of histological data that were used as a reference standard to validate imaging findings.

PET/MRI Findings
An example of whole-body biodistribution of 68 Ga-PSMA PET and 68 Ga-DOTA-RM2 PET is reported in Figure 1. Physiological high 68 Ga-PSMA uptake can be visualised in the salivary and lacrimal glands, liver, spleen, small intestine, kidneys, urinary bladder and ureters ( Figure 1A), while 68 Ga-DOTA-RM2 showed physiological high uptake in the pancreatic gland and urinary bladder ( Figure 1B). 68 Ga-PSMA PET detected intra-prostatic lesions in all patients, while 68 Ga-DOTA-RM2 PET identified the intraprostatic disease in 18/19 patients. Additionally, in 2/22 patients 68 Ga-PSMA PET also detected seminal vesicles uptake. The specific sites of intra-prostatic 68 Ga-PSMA and 68 Ga-DOTA-RM2 uptake are reported in Table 2.    Table 2.   Table 3.   Table 4. TB 68 Ga-PSMA images revealed a suspicion for lymph nodal involvement in 7/22 patients, and for bone involvement in 3/22. In TB 68 Ga-DOTA-RM2 PET images a pathological focal uptake was detected at lymph nodal level in 4/19 patients with no evidence of bone metastases in any patient. The detailed description of sites of lymph nodal and bone 68 Ga-PSMA and 68 Ga-DOTA-RM2 uptake are reported in Table 2.

Comparisons between 68 Ga-PSMA PET, 68 Ga-DOTA-RM2 PET and MRI and Validation with Histology
Regarding intraprostatic disease, in 16/19 patients the site of the primary prostatic lesion was concordant among the three imaging modalities (see as an example, patient n. 10, Figure 2). ADC: Apparent Diffusion Coefficient; Volume: Manually segmented volume on 3D Slicer.

Comparisons between 68 Ga-PSMA PET, 68 Ga-DOTA-RM2 PET and MRI and Validation with Histology
Regarding intraprostatic disease, in 16/19 patients the site of the primary prostatic lesion was concordant among the three imaging modalities (see as an example, patient n. 10, Figure 2). Histological examination was available for 11 of these patients, and, whenever present, confirmed these findings. In the three patients who did not undergo 68 Ga-DOTA-RM2, 68 Ga-PSMA and MRI were concordant in identifying the intraprostatic pathological findings (n. 19, n. 20, n. 21, Table 2).
Among the patients for whom discordant imaging findings were observed, in 1/19 MRI and 68 Ga-DOTA-RM2 detected bilateral pathological findings, with 68 Ga-PSMA showing radiotracer uptake only in correspondence of the left lobe (patient n. 16, Table 2). In 1/19 patient (n. 18, Table 2) MRI identified two pathological findings in the right and Histological examination was available for 11 of these patients, and, whenever present, confirmed these findings. In the three patients who did not undergo 68 Ga-DOTA-RM2, 68 Ga-PSMA and MRI were concordant in identifying the intraprostatic pathological findings (n. 19, n. 20, n. 21, Table 2).
Among the patients for whom discordant imaging findings were observed, in 1/19 MRI and 68 Ga-DOTA-RM2 detected bilateral pathological findings, with 68 Ga-PSMA showing radiotracer uptake only in correspondence of the left lobe (patient n. 16, Table 2). In 1/19 patient (n. 18, Table 2) MRI identified two pathological findings in the right and left side of the prostate, respectively, showing 68 Ga-PSMA uptake in correspondence of the right lobe and a negative 68 Ga-DOTA-RM2 PET. These patients have not undergone radical prostatectomy yet, therefore histological examination was not yet available to validate these findings.
Finally, in 1/19, (n. 15, Table 2), 68 Ga-DOTA-RM2 PET and MRI were concordant in identifying a pathological finding in the right side of the prostate, while 68G a-PSMA PET showed a focal uptake in the left lobe; histological examination demonstrated a bilateral prostate cancer with the dominant neoplastic nodule being located in the right lobe.
In terms of the local extension, SVI was detected by MRI in seven patients (n. 3, n. 5, n. 16, n. 18, n. 20, n. 21 and n. 22, Table 2), by 68 Ga-PSMA in two patients (n. 3, n. 20, Table 2), while no uptake was present on 68 Ga-DOTA-RM2 images. No histological examination was available for these patients to confirm the imaging findings.
Histological validation was available for patients n. 1, n. 7 and n. 17, confirming the presence of left external iliac nodal metastases in patient n. 1 and absence of nodal metastases in patients n. 7 and n. 17. Moreover, the lymphnodal involvement described on 68 Ga-PSMA PET in correspondence of the perivesical fat for patient n. 1 was not confirmed as these lymphnodes were not removed during surgery. Venn diagrams showing the true positive, false positive, true negative and false-negative findings for lymph nodes involvement validated by means of histological specimens for all investigated imaging modalities are depicted in Figure 4. 2), while no uptake was present on 68 Ga-DOTA-RM2 images. No histological examination was available for these patients to confirm the imaging findings. MRI identified ECE in 10 patients (n. 3, n. 5, n. 6, n. 8, n. 10, n. 11, n. 12, n. 14, n. 16, n. 18, Table 2); among the 4/10 patients with the availability of histological confirmation, ECE was confirmed in only 2/4 patients (n. 8, n. 12, Table 2). Both 68 Ga-PSMA and 68 Ga-DOTA-RM2 PET are not suitable to identify ECE of PCa, because of the limited spatial resolution compared to MRI.

Figure 4.
Venn diagram showing the true-positive, false-positive, true-negative and false-negative findings regarding lymph node involvement for all the investigated imaging modalities using histopathological specimens acquired during radical prostatectomy as ground truth.
Regarding distant metastases, 68 Ga-PSMA showed increased pathological uptake at a bone level in three patients (n. 2, n. 6, n. 18, Table 2), 68 Ga-DOTA-RM2 did not detect any pathological uptake at bone level. (Figure 5). DICE score was computed to quantitatively assess the overlap between the volume of the primary intra-prostatic lesions manually segmented on 68 Ga-PSMA PET, 68 Ga- DICE score was computed to quantitatively assess the overlap between the volume of the primary intra-prostatic lesions manually segmented on 68 Ga-PSMA PET, 68 Ga-DOTA-RM2 PET and MR images at the individual level. On average, the DICE score between 68 Ga-PSMA and MRI = 0.51(range: 0.00-0.79); between 68 Ga-PSMA and 68 Ga-DOTA-RM2 = 0.41 (range: 0.05-0.72); while the DICE score between 68 Ga-DOTA-RM2 and MRI = 0.36 (range: 0.07-0.72). DICE scores for each patient across the investigated modalities are reported in Table 6 ( Figure 6).

Correlations between PET Semi-Quantitative and MRI Quantitative Imaging Parameters
None of the investigated semi-quantitative 68 Ga-PSMA PET parameters significantly correlated with its correspondent parameter on 68 Ga-DOTA-RM2 PET images, however, the volume of the primary tumour manually segmented on 68 Ga-PSMA PET was highly correlated with the one manually contoured on MR images (rho = 0.697, p = 0.003). MRI quantitative parameters did not correlate with 68 Ga-DOTA-RM2 PET semi-quantitative parameters.
Tumour volume manually segmented on MR images presented a moderate association with GS that approached the level of significance (rho = 0.419, p = 0.053), while 68 Ga-PSMA PET and 68 Ga-DOTA-RM2 semi-quantitative parameters did not correlate with any of the considered clinical data (p-value ≥ 0.05).

Discussion
The present pilot study reports our preliminary experience on the use of 68 Ga-PSMA and 68 Ga-DOTA-RM2 PET/MRI imaging in high-risk prostate cancer staging.
Few studies have investigated prostate cancer by using both 68 Ga-PSMA and 68 Ga-DOTA-RM2 PET so far, both in the staging [15] and restaging setting of the disease [18,19].
In our cohort of patients, differently from all the other published papers, all subjects were studied by using a hybrid PET/MRI scanner both for 68G a-PSMA and 68 Ga-DOTA-RM2 radiotracers [15,18,19].
In fact, among the few published studies that investigated the role of this peculiar multitracer approach in PCa, PET/MRI and PET/CT were used alternatively for 68 Ga-PSMA and 68 Ga-DOTA-RM2 PET scans [18,19] or PET/CT were adopted as the only hybrid imaging modality [15].
In the setting of PCa staging, Schollhammer and colleagues reported a clinical case of a patient undergoing PET/CT scans with 68 Ga-PSMA, 68 Ga-RM2 and 18F-Choline, while Fassbender et al. used 68 Ga-PSMA PET/CT and 68 Ga-Ga-RM2 PET/MRI to study eight patients with a primary diagnosis of PCa [15,24]. The same heterogeneity in terms of the type of scanners used for patients' scanning can be also observed in the few studies assessing the role of 68 Ga-PSMA and 68 Ga-DOTA-RM2 in patients with recurrent PCa. The first study performing a comparative evaluation between these two radiotracers in recurrent PCa is the one by Minamimoto et al. In this pioneering work, comparing the biodistribution of 68 Ga-PSMA-11 and 68 Ga-RM2 in a small cohort of patients with biochemically recurrent PCa, PET/CT was adopted for 68 Ga-PSMA studies while PET/MRI scanner was used for 68 Ga-DOTA-RM2 PET acquisitions [18].
Similarly, Baratto et al. recently published a study on the use of 68 Ga-PSMA and 68 Ga-DOTA-RM2 in a cohort of patients with recurrent PCa and compared the diagnostic performances of these two radiotracers. They showed that 68 Ga-PSMA11 and 18F-DCFPyL might have a complementary role as they detect different sites of disease recurrence. Notably, the group used a PET/MR scanner only for 68 Ga-RM2 imaging and regarding PSMA PET/CT scans, 68 Ga-PSMA11 or 18F-DCFPyL were alternatively used [19].
The use of a PET/MRI scanner in the staging phase of PCa allows to perform a diagnostic MRI on the pelvic region, thus obtaining all the necessary morphological and multiparametric information for accurate identification and characterisation of the primary tumour. Moreover, the possibility to simultaneously acquire a PET scan with two different radiotracers assessing different metabolic pathways provides additional information regarding primary tumour characteristics, together with a whole-body evaluation of the disease. Finally, the use of a PET/MRI scanner rather than PET/CT scanner strongly reduces the radiation exposure for the patient [25].
Differently from other groups that investigated the dual tracer approach of 68 Ga-PSMA and 68 Ga-DOTA-RM2 in PCa staging, or restaging, using a PET/CT scanner [15,18,19], one of the most relevant patients' advantages in the present study relies on the possibility to have received a diagnostic MRI simultaneously acquired to the PET image acquisition. In fact, MRI is expected to increase the diagnostic accuracy of PET imaging for local staging (ECE and SVI) [26], and the information derived from both modalities could be incorporated into clinical nomograms to significantly enhance the pre-operative staging accuracy [27,28]. Moreover, MRI shows excellent diagnostic performance in the detection of bone metastases [29]. If used in combination with PET, MRI could provide complementary information on the bone disease when PET findings are equivocal or when metastatic lesions do not show significant PSMA uptake. Finally, WB-MRI could be of added value in monitoring the response to loco-regional or systemic treatments [30,31].
In our cohort of patients, the primary PCa was detected in 18/19 patients by all three imaging modalities. In our cohort of patients, histological examination was available only for a minority of subjects and, regarding the intraprostatic disease, was used as the gold standard to confirm the findings detected by the three imaging modalities. In one patient (n. 15) MRI and 68 Ga-DOTA-RM2 identified a lesion in the right prostatic lobe, with 68 Ga-PSMA being positive in the left lobe. Histological examination reported bilateral prostatic neoplasia with a right dominant nodule. The discrepancies observed between 68 Ga-PSMA and 68 Ga-RM2 might reflect a complementary role of these imaging modalities in identifying different intraprostatic foci of disease presenting different metabolic patterns, thus enlightening a synergic role of 68 Ga-PSMA and 68 Ga-DOTA-RM2 in prostate cancer, in line with previously published data [15,18,19].
For instance, Fassbender et al. in their cohort of eight patients with primary PCa undergoing 68G a-PSMA PET/CT and 68 Ga-Ga-RM2 PET/MRI concluded that the qualitative findings of PET scans could provide combined relevant information. In their study, both radiotracers partially showed the same tumour region and, in some cases, different tumour parts, thus providing a better PCa characterisation and reflecting the heterogeneous and sometimes polyclonal behaviour that characterise PCa [15]. Similarly, the results reported by Baratto and colleagues, comparing RM2-PET and PSMA-PET in patients with biochemically recurrent PCa and by Iagaru and colleagues in patients with newly diagnosed intermediate-or high-risk prostate cancer suggested that the use of both 68 Ga-PSMA and 68 Ga-DOTA-RM2 provided different and complementary information on PCa [19,32].
To investigate the correspondence of the intra-prostatic findings referable to the site of the primary tumour across modalities, DICE score between manually segmented primary intra-prostatic tumour volumes on 68 Ga-PSMA, 68 Ga-DOTA-RM2 PET and MR images were calculated. Volumes of the primary tumours, as defined in all the investigated imaging modalities largely overlap. The highest mean DICE score was the one between 68 Ga-PSMA PET and MRI. This may be partially explained by the fact that 68 Ga-DOTA-RM2 tumour volumes were generally smaller, and partially by the fact that 68 Ga-PSMA PET and MRI were simultaneously acquired, thus being intrinsically co-registered. Conversely, an automatic co-registration tool on 3D Slicer, with manual adjustments when needed, was used to overlap 68 Ga-DOTA-RM2 PET images to 68 Ga-PSMA and MR diagnostic images. A residual component of noise might have hampered the computation of the DICE score, therefore resulting in a minor overlap between 68 Ga-DOTA-RM2 PET and the other images.
Spearman correlations between multitracer PET and MRI parameters revealed a significant, strong, correlation between tumour volume manually segmented on 68 Ga-PSMA PET and MR images. No significant associations between parameters derived from PET with different radiotracers were found. This is in contrast with the strong association between 68 Ga-PSMA PET and 68 Ga-DOTA-RM2 PET SUVmax and SUVmean reported by Minamimoto et al. in 2016 [18]. However, that pioneering work relied on a very small sample (n = 7) of patients presenting with biochemical recurrence. Future studies with larger cohorts of patients are needed to unravel the possible association between semiquantitative parameters derived from 68 Ga-PSMA and 68 Ga-DOTA-RM2 PET. Furthermore, concerning the correlation between imaging parameters and clinical data, a moderate correlation between tumour volume manually segmented on MR images and GS at biopsy that approached the level of significance (p = 0.053) was detected. All the other tested correlations resulted non-significant and this is in line with what is reported by Fassbender et al. [15]. These results have to be interpreted with great caution and more evidence is needed before speculating on the clinical utility of these findings since our sample was small and quantitative analyses might be susceptible to lack of statistical power. Future studies, with larger samples, will allow for the unraveling of the possible association between semi-quantitative PET, quantitative MRI parameters and clinical data.
Some limitations should be pointed out regarding the present study. First of all, histopathological correlation with the post-surgical specimen was available only for those patients who have undergone radical prostatectomy. Therefore, being the present analysis a pilot study with only a preliminary evaluation of collected imaging and histological data, the analysis will certainly be improved as soon as all histological data are available, with also a detailed co-registration between imaging and histopathological data as well as correlations between imaging findings and histological examinations. Another limitation of this study is the low number of patients. However, besides the fact that the few papers already published on PCa staging and using both 68 Ga-PSMA and 68 Ga-RM2 PET radiotracers included a number of patients even lower than the one presented in the present paper [15], we consider that these preliminary data are interesting to underlie the potential complementary and synergic role of the two different PET radiotracers together with mp-MRI.

Conclusions
Based on the results of the present study, a potential complementary role of 68 Ga-PSMA and 68 Ga-DOTA-RM2 in PCa staging can be enlightened, in view of the different findings detected by the two imaging modalities in some of the patients included in our cohort. In fact, the possibility to identify different sites of disease by using a multitracer approach certainly improves the disease characterization and therefore it may ultimately have an impact on patients' management and follow-up. These findings should be validated on larger cohorts of patients to definitively assess the utility of hybrid 68 Ga-PSMA PET/MRI and 68 Ga-DOTA-RM2 PET/MRI in the clinical management of PCa.