This was a prospective study in PCa patients with suspected bone metastases to evaluate the performance of PSMA PET/CT compared to fluoride PET/CT. The results suggested that PSMA PET/CT was able to detect most of the bone lesions (83%) that were positive on fluoride PET/CT. Furthermore, moderate but significant correlation of PSA at time of scan with number of bone metastases and SUVmax supports the use of PSMA PET/CT as the most optimal imaging tool for restaging of PCa.
The usefulness of PSMA PET/CT has primarily been investigated with a focus on localizing biochemical relapse of PCa [21, 25, 28]. A number of studies (summarized in supplementary table 1) have investigated the diagnostic accuracy of PSMA PET/CT regarding bone metastasis compared to BS [29, 30]. In addition, few studies compared the diagnostic performance of several clinically available imaging modalities in localizing bone spread in PCa patients [31–37]. These studies showed better diagnostic performance of fluoride PET/CT and PSMA PET/CT compared to other modalities such as BS, SPECT/CT, WB-MRI, and choline PET/CT [see supplementary Table 1 for details]. Further, PSMA PET/CT showed additional value with improved specificity but with an overlapping sensitivity in comparison to fluoride PET/CT. The results from this study on comparison of PSMA PET/CT with fluoride PET/CT were also in line with these studies.
One important finding from this study is a higher detection rate of fluoride PET/CT compared to -PSMA PET/CT (699 vs 579 bone lesions). In comparison to our study, a retrospective study with a smaller data set (n = 16) conducted by Uprimny et al. also documented higher detection rate of fluoride PET/CT . In that study, the authors observed low uptake of PSMA in osteosclerotic lesions similar to , stated as the possible explanation for low detection rate of bone lesions on PSMA PET/CT. In concordance, we also noticed overall low intensity of PSMA uptake in sclerotic bone lesions (Fig. 4). However, PSMA-PET had the influential role with its ability to detect non-osseous PCa spread. In seven patients without bone disease on both PET scans, the presence of lymph node lesions on PSMA-PET/CT changed the treatment decision.
Determining the presence of oligo-metastatic bone lesions, the potential targets for metastatic-directed therapies are clinically relevant as they can be irradiated. In the six patients who were identified with oligo-metastatic bone lesions with both PSMA and fluoride PET/CT, one patient showed additional non-osseous lesions on PSMA PET/CT, altering the treatment plan. The remaining five subjects received radiation therapy. Three more patients were identified with oligo-metastatic bone status on PSMA PET/CT but having more than five lesions on fluoride PET/CT. In addition to bone lesions, PSMA PET/CT also showed either local relapse or lymph node lesions in these patients which influenced the treatment management. However, this cohort is not large enough to determine the added value of PSMA PET/CT treatment decisions related to oligometastatic disease.
Several prospective studies showed that the detection rate of PSMA PET/CT depends on PSA levels but, due to occasionally very high receptor expression, small lesions can sometimes be visualised even at PSA levels as low as 0.2 ng/mL [36, 39]. Concordantly, our study with focus on skeletal metastasis also showed that number of bone lesions on PSMA PET/CT is associated with PSA levels at time of scan (r = 0.45, p = 0.02). Furthermore, significant difference in PSA levels between patients with positive and negative bone findings (64.4 ng/mL vs 10.0 ng/mL, p = 0.03) supports the association of PSMA PET/CT positivity rate and PSA. Despite the strong correlation of tumour volume and disease burden between PSMA and fluoride PET/CT, none of these metrics correlated with PSA. PSA did not predict tumour burden in the given cohort. However, strong correlation of tumour burden with fluoride PET/CT favours the use of PSMA PET/CT.
The choice of imaging techniques for restaging of PCa at a given centre depends on local parameters such as cost-effectiveness, accessibility and expertise. Many hospitals in Sweden still use BS along with an abdominal CE-CT scan as standard in the diagnostic workup. Fluoride PET/CT is generally considered superior to 99mTc-MDP-BS and 99mTc-MDP-SPECT/CT for detection of bone metastasis [35, 40, 41]. In further support, few studies evaluated the impact of fluoride PET/CT on patient prognosis [42, 43]. At our hospital, fluoride PET/CT is recommended over BS in PCa re-staging to locate early signs of bone disease and is generally performed with CE-CT to detect soft tissue lesions.
Using follow-up scanning we could show that some lesions are detected earlier with fluoride PET/CT than with PSMA PET/CT. In none of these cases did the higher sensitivity of fluoride PET/CT lead to therapy changes and the measured tumour burden was similar for both tracers. However, equivocal skeletal findings on fluoride PET/CT are relatively common and contribute to false positive cases [32, 35], which might require additional diagnostic procedures. In this study, we also found equivocal bones lesions on PSMA PET/CT, but the overall perception is that this is a smaller problem with PSMA than with fluoride PET/CT. In addition, PSMA PET/CT provided additional information in detecting local recurrences and lymph node metastases, thus influencing the management, as seen in 7 subjects in our current study. Further, PSMA-PET can be used to select patients with potential benefit from targeted radioligand therapies using 177Lu  and 225Ac.
This study has several limitations. It was based on a small selective group of patients with high suspicion for widespread disease involving bone. A standard reference, preferably histological reports, to confirm the positive findings of PET imaging is missing. However, accessing bone for biopsy collection is neither ethically nor practically possible in all lesions. Follow-up scans with optimal imaging modalities were not available in all patients.