Simplified Methods for Quantification of ¹⁸F-DCFPyL Uptake in Patiens with Prostate Cancer

Objectives: Radiolabeled Prostate-Specific Membrane Antigen Positron Emission Tomography (PSMA PET) has demonstrated favourable diagnostic results for prostate cancer (PCa) imaging. Quantification of PSMA radiotracer uptake is desired to foster reliable interpretation of PET-images, use PSMA uptake as an imaging biomarker, and monitor treatment effects. The aim of this study was to perform a full pharmacokinetic analysis of ¹⁸F-DCFPyL, a 2^nd generation ¹⁸F-labelled PSMA ligand. Based on the pharmacokinetic analysis (reference method), simplified methods for quantification of ¹⁸F-DCFPyL uptake were validated for use in routine clinical practice.

Methods: Eight patients with metastasized PCa were included. Dynamic PET acquisitions were performed on a Philips Ingenuity TF PET/CT scanner from 0-60 and 90-120 min, starting after injection of a median dose of 313MBq ¹⁸F-DCFPyL (range 292-314 MBq). Continous and manual arterial blood sampling provided calibrated plasma tracer input funtions. Time-activity curves were derived for each PCa lesion. Several tissue-compartment models were fitted per PCa lesion and the optimal model for ¹⁸F-DCFPyL kinetics was selected based on the Akaike criterion. Simplified methods for quantification of ¹⁸F-DCFPyL uptake (e.g. Standardized Uptake Values, SUV; Tumor-to-Blood Ratios, TBR) were validated against the reference kinetic parameter derived from full pharmacokinetic analysis.

Results: N=46 metastatic lesions were evaluated. No metabolites of ¹⁸F-DCFPyL were detected in the blood samples. Tumor activity concentrations continued to rise over the course of the PET study (up until 120 min post injection). A reversible two-tissue compartment model was the preferred model for ¹⁸F-DCFPyL kinetics in 59% of lesions. However, the observed k₄ was rather small, resulting in nearly irreversible kinetics during the course of the PET study. Hence, k₄ was fixated at 0.015 and K_i was preferred as the reference pharmacokinetic parameter. Whole-blood TBR provided excellent correlation with K_i derived from full kinetic analyis (R² = 0.97). These TBR could be simplifed further by replacing the blood samples with a single image-based measurement of blood activity in the ascending aorta (image-based TBR, R­² = 0.96). SUV correlated poorly with K_i (R­² = 0.47 and R­² = 0.60 for SUV normalized to body weight and lean-body mass, respectively), which is most likely due to deviant blood activity concentrations (i.e. tumor tracer input) in patients with higher tumor volumes.

Conclusions: ¹⁸F-DCFPyL kinetics in PCa lesions are best described by a reversible two-tissue compartment model with a fixed value for k₄. Scanning at later time points after radiotracer injection (e.g. 120 minutes) results in higher tumor activity concentrations. Image-based Tumor-to-Blood ratios were validated as a simplified method to quantify ¹⁸F-DCFPyL uptake and can be applied to clinical, whole-body PET scans. SUV did not provide reliable quantification of ¹⁸F-DCFPyL uptake due to variations in the tracer input function across subjects.

• Download figure
• Open in new tab
• Download powerpoint