Rationale: In prostate cancer (PCa) patients, the Tumor-to-Blood ratio (TBR) has been validated as the preferred simplified method for lesional 18F-DCFPyL (a radiolabeled Prostate-Specific Membrane Antigen (PSMA) ligand) uptake quantification on positron emission tomography (PET). In contrast to standardized uptake values (SUV), the TBR accounts for variability in arterial input functions caused by differences in total tumor-burden between patients (the 'sink effect'). However, TBR depends strongly on tracer uptake interval, has worse repeatability and is less applicable in clinical practice than SUVs. We investigated whether SUV could provide adequate quantification of [18F]DCFPyL uptake on PET/computed tomography (CT) in a patient cohort with low prostate cancer (PCa) burden. Methods: A total of 116 patients with PCa undergoing 18F-DCFPyL PET/CT imaging were retrospectively included. All 18F-DCFPyL-avid lesions suspect for PCa were semi-automatically delineated. SUVpeak was plotted against TBR for the most intense lesion of each patient. The correlation of SUVpeak and TBR was evaluated using linear regression, and was stratified for patients undergoing PET/CT for primary staging, restaging at biochemical recurrence and in metastatic castration-resistant PCa. Moreover, the correlation was evaluated as a function of tracer uptake time, Prostate-Specific Antigen (PSA)levels and PET-positive tumor volume. Results: A total of 436 lesions was delineated (median 1 per patient, range 1-66). SUVPeak correlated well to TBR in patients with PCa and a total tumor volume of <200 ml (R2=0.931) (Figure 1). The correlation between SUV and TBR was not affected by disease setting, PSA levels or tumor volume. SUVpeak depended less on tracer uptake time than TBR. Conclusion: For 18F-DCFPyL PET/CT, SUVpeak highly correlates with TBR. Therefore, it is a valuable simplified semi-quantitative measurement in patients with low volume prostate cancer (<200 ml). SUVpeak can therefore be applied in 18F-DCFPyL PET assessment as an imaging biomarker to characterize tumors and to monitor treatment outcomes.
Journal of nuclear medicine : official publication, Society of Nuclear Medicine. 2021 Jan 28 [Epub ahead of print]
Yves Jan Lambert Bodar, Berend Pieter Francis Koene, Bernard H E Jansen, Matthijs C F Cysouw, Dennie Meijer, Harry N H Hendrikse, André N Vis, Ronald Boellaard, Daniela E Oprea-Lager
Amsterdam UMC, location VUmc, Netherlands.