Is there more than meets the eye in PSMA imaging in prostate cancer with PET/MRI? Looking closer at uptake time, correlation with PSA and Gleason score

Borjana Bogdanovic; Esteban L Solari; Alberto Villagran Asiares; Sandra van Marwick; Sylvia Schachoff; Matthias Eiber; Wolfgang A Weber; Stephan G Nekolla

doi:10.1186/s41824-023-00166-5

Is there more than meets the eye in PSMA imaging in prostate cancer with PET/MRI? Looking closer at uptake time, correlation with PSA and Gleason score

Eur J Hybrid Imaging. 2023 Apr 17;7(1):8. doi: 10.1186/s41824-023-00166-5.

Authors

Borjana Bogdanovic¹, Esteban L Solari², Alberto Villagran Asiares², Sandra van Marwick², Sylvia Schachoff², Matthias Eiber², Wolfgang A Weber², Stephan G Nekolla²

Affiliations

¹ Department of Nuclear Medicine, School of Medicine, Klinikum Rechts Der Isar, Technical University of Munich, Ismaninger Str. 22, 81675, Munich, Germany. bogdanovicborjana@gmail.com.
² Department of Nuclear Medicine, School of Medicine, Klinikum Rechts Der Isar, Technical University of Munich, Ismaninger Str. 22, 81675, Munich, Germany.

Abstract

Background: In patients with increasing PSA and suspicion for prostate cancer, but previous negative biopsies, PET/MRI is used to test for tumours and target potential following biopsy. We aimed to determine different PSMA PET timing effects on signal kinetics and test its correlation with the patients' PSA and Gleason scores (GS).

Methods: A total of 100 patients were examined for 900 s using PET/MRI approximately 1-2 h p.i. depending on the tracer used (⁶⁸Ga-PSMA-11, ¹⁸F-PSMA-1007 or ¹⁸F-rhPSMA7). The scans were reconstructed in static and dynamic mode (6 equal frames capturing "late" PSMA dynamics). TACs were computed for detected lesions as well as linear regression plots against time for static (SUV) and dynamic (SUV, SUL, and percent injected dose per gram) parameters. All computed trends were tested for correlation with PSA and GS.

Results: Static and dynamic scans allowed unchanged lesion detection despite the difference in statistics. For all tracers, the lesions in the pelvic lymph nodes and bones had a mostly negative activity concentration trend (78% and 68%, resp.), while a mostly positive, stronger trend was found for the lesions in the prostate and prostatic fossa following RPE (84% and 83%, resp.). In case of ⁶⁸Ga-PSMA-11, a strong negative (R_min = - 0.62, R_max = - 0.73) correlation was found between the dynamic parameters and the PSA. ¹⁸F-PSMA-1007 dynamic data showed no correlation with PSA, while for ¹⁸F-rhPSMA7 dynamic data, it was consistently low positive (R_min = 0.29, R_max = 0.33). All tracers showed only moderate correlation against GS (R_min = 0.41, R_max = 0.48). The static parameters showed weak correlation with PSA (R_min = 0.24, R_max = 0.36) and no correlation with GS.

Conclusion: "Late" dynamic PSMA data provided additional insight into the PSMA kinetics. While a stable moderate correlation was found between the PSMA kinetics in pelvic lesions and GS, a significantly variable correlation with the PSA values was shown depending on the radiotracer used, the highest being consistently for ⁶⁸Ga-PSMA-11. We reason that with such late dynamics, the PSMA kinetics are relatively stable and imaging could even take place at earlier time points as is now in the clinical routine.

Keywords: Late dynamic imaging; MR; PET quantification; PSMA; Prostate cancer; Whole-body PET.