From 18 F-choline to prostate-specific membrane antigen PET: lessons from early multiparametric kinetic modelling in prostate cancer.

BACKGROUND

We revisited our early multiparametric dynamic choline PET data to demonstrate methodological feasibility for kinetic modelling and to highlight lessons relevant for current prostate-specific membrane antigen (PSMA) PET protocols.

METHODS

Nine men (aged 48-81 years) with biochemical recurrence of prostate cancer (prostate-specific antigen [PSA] range: 0.12-45.1 μg/L) underwent dynamic 18 F-FECH PET/computed tomography. Patlak analysis produced two outcome values, influx rate ( Ki ) and volume of distribution. For quantitative analysis region of interests (ROIs) were drawn on a metabolically active focus of the prostate. For semiquantitative analysis the tracer accumulation in the ROIs was measured using the standardised uptake value. Arterial input functions were derived visually, and Patlak graphical analysis was performed. Correlations were explored between kinetic parameters, Gleason grade, and serum PSA. Spearman's correlation coefficients were calculated to assess the possible association between 18 F choline uptake, clinical and histopathological characteristics.

RESULTS

Dynamic acquisitions were technically feasible and analyzable before urinary excretion obscured the prostate. Time-activity curves plateaued after 1-2 min, and Patlak plots showed bilinear behaviour with measurable Ki and volume of distribution. Cancerous prostate tissue demonstrated increased perfusion compared with background tissue. No significant correlation was observed between kinetic parameters and PSA or Gleason grade.

CONCLUSION

Dynamic 18 F-choline PET provided quantitative kinetic parameters reflective of tumour perfusion. Although choline imaging has been largely superseded by PSMA PET because of superior tumour-to-background contrast and specificity, the technical principles established in this study remain directly applicable to current PSMA kinetic and parametric imaging.