LAFOV PSMA PET/CT: Preliminary Report on Protocol Optimization and Future Perspectives in Prostate Cancer Imaging
摘要
Long axial field-of-view (LAFOV) PET/CT systems provide up to 40-fold higher sensitivity than standard digital systems. This study evaluated the impact of reducing acquisition times in [18 F]DCFPyL (PSMA) total-body PET/CT to support future research on the development of low-dose protocols.
MethodsEighty-five patients with prostate cancer underwent total-body PET/CT on a 194 cm-LAFOV system. Standard 8-minute list-mode acquisitions were retrospectively reconstructed into 6-, 4-, and 2-minute datasets. In a subgroup of thirty-three consecutive patients, two experts independently assessed image quality (5-point Likert scale) and E-PSMA criteria. Quantitative analysis (SUVmax, SUVmean) was performed for primary, nodal, and bone lesions.
ResultsIn the whole dataset (n = 85) the lesion detection rate (90.5%) and classification consistency between readers and across acquisition times exceeded 95% and remained stable between the 8-minute and 2-minute reconstructions. In a sub-group of thirty-three, no significant differences were found in quantitative parameters across all acquisition times (p > 0.05). While image quality scores improved slightly with longer durations (ranging from 4.50 at 2 min to 4.82 at 8 min), lesion visibility remained high even at 2 min. Changes in physiological background uptake did not impact lesion quantification.
ConclusionThe preliminary results of the study suggest that LAFOV-PSMA-PET/CT preserves quantitative accuracy and lesion detectability even with acquisition times reduced to 2–4 min. These findings demonstrate that high-sensitivity systems might significantly shorten protocols without compromising diagnostic integrity, providing a potential rationale for further optimization of imaging protocols and for future prospective studies specifically designed to evaluate whether the administered radiotracer activity can be safely reduced to ultra-low-dose levels while maintaining diagnostic performance.