Quantification of [¹⁸F]fluorocholine in prostate cancer

Objectives Choline kinase is upregulated in prostate cancer (PCa), resulting in increased [¹⁸F]fluorocholine ([¹⁸F]FCH) uptake. The purpose of this study was to identify the best kinetic model for quantifying [¹⁸F]FCH uptake and to validate simplified methods for routine clinical use.

Methods 8 PCa patients with metastatic lymph nodes (LN) underwent a 40 min dynamic PET scan after bolus injection of 200 MBq [¹⁸F]FCH. Metabolite corrected plasma input functions were obtained using continuous arterial blood sampling and 6 manual arterial blood samples for plasma-to-whole blood ratio and radiolabeled metabolite information. In addition, 3 manual venous blood samples were obtained. LN time-activity curves were fitted to several single and two tissue compartment models with and without blood volume parameter. The optimal model was selected based on both Akaike Information Criterion and robustness of the kinetic parameters. Next, several standardized uptake values (SUV) and SUV_AUC, defined as LN activity concentration at 35-40 min normalized to the area under the curve (AUC) of the plasma input function over 0-30 min, were compared with pharmacokinetic parameters obtained using the optimal model.

Results Parent fractions were 0.12±0.4 after 20 min, necessitating individual metabolite corrections. Best fits were obtained using an irreversible single tissue compartment model with blood volume parameter (1T1kV_B). Bland-Altman analysis showed high variability in parent fraction (limits of agreement 63%) and 4% underestimation in plasma-to-whole blood ratios from venous versus arterial blood samples. SUV showed poor correlation with K₁ (R²<0.27). SUV_AUC yielded higher correlations (R²=0.83; R²=0.66 if based on whole-blood AUC).

Conclusions [¹⁸F]FCH uptake can be quantified using a 1T1kV_B with arterial blood sampling. SUV is not reliable. A clinical compromise could be SUV_AUC derived from two consecutive static PET scans, one centered over a large blood pool structure during 0-30 min p.i. to obtain whole-blood AUC, the other a whole body scan at 35 min p.i. to obtain LN activity concentrations.