A Retrospective Comparative Study of Sodium Fluoride Na¹⁸F-PET/CT and ⁶⁸Ga-PSMA-11 PET/CT in the Bone Metastases of Prostate Cancer Using a Volumetric 3-D Radiomic Analysis

Bone is the most common metastatic site in prostate cancer (PCa). 68Ga-PSMA-11 (or gozetotide) and sodium fluoride-18 (Na18F) are rather new radiopharmaceuticals for assessing PCa-associated bone metastases. Gozetotide uptake reflects cell membrane enzyme activity and the sodium fluoride uptake measures bone mineralization in advanced PCa. Here, we aim to characterize this difference and possibly provide a new method for patient selection in targeted therapies. Methods: The study consisted of 14 patients with advanced PCa (M group > 5 lesions), who had had routine PET/CT both with PSMA and NaF over consecutive days, and 12 PCa patients with no skeletal metastases (N). The bone regions in CT were used to coregister the two PET/CT scans. The whole skeleton volume(s) of interest (VOIs) were defined using the CT component of PET (HU > 150); similarly, the sclerotic/dense bone was defined as HU > 600. Additional VOIs were defined for PET, with pathological threshold values for PSMA (SUV > 3.0) and NaF (SUV > 10). Besides the pathological bone volumes measured with each technique (CT, NaF, and PSMA-PET) and their contemporaneous combinations, overlapping VOIs with the CT-based skeletal and sclerotic volumes were also recorded. Additionally, thresholds of 4.0, 6.0, and 10.0 were tested for SUVPSMA. Results: In group M, the skeletal VOI volumes were 8.77 ± 1.80 L, and the sclerotic bone volumes were 1.32 ± 0.50 L; in contrast, in group N, they were 8.73 ± 1.43 L (skeletal) and 1.23 ± 0.28 L (sclerosis). The total enzyme activity for PSMA was 2.21 ± 5.15 in the M group and 0.078 ± 0.053 in the N group (p < 0.0002). The total bone demineralization activity for NaF varied from 4.31 ± 6.17 in the M group and 0.24 ± 0.56 in group N (p < 0.0002). The pathological PSMA volume represented 0.44–132% of the sclerotic bone volume in group M and 0.55–2.3% in group N. The pathological NaF volume in those patients with multiple metastases represented 0.27–68% of the sclerotic bone volume, and in the control group, only 0.00–6.5% of the sclerotic bone volume (p < 0.0003). Conclusions: These results confirm our earlier findings that CT alone does not suit the evaluation of the extent of active skeletal metastases in PCa. PSMA and NaF images give complementary information about the extent of the active skeletal disease, which has a clinical impact and may change its management. The PSMA and NaF absolute volumes could be used for planning targeted therapies. A cut-off value 3.0 for SUVPSMA given here is the best correlation in the presentation of active metastatic skeletal disease.