Kinetic Analysis and Metabolism of Poly(Adenosine Diphosphate-Ribose) Polymerase-1-Targeted 18F-Fluorthanatrace PET in Breast Cancer

Abstract

The poly(adenosine diphosphate-ribose) polymerase inhibitors (PARPi) have demonstrated efficacy in ovarian, breast, and prostate cancers, but current biomarkers do not consistently predict clinical benefit. ¹⁸F-fluorthanatrace (¹⁸F-FTT) is an analog to rucaparib, a clinically approved PARPi, and is a candidate biomarker for PARPi response. This study intends to characterize ¹⁸F-FTT pharmacokinetics in breast cancer and optimize image timing for clinical trials. A secondary aim is to determine whether ¹⁸F-FTT uptake in breast cancer correlates with matched frozen surgical specimens as a reference standard for PARP-1 protein. Methods: Thirty prospectively enrolled women with a new diagnosis of breast cancer were injected with ¹⁸F-FTT and imaged dynamically 0-60 min after injection over the chest, with an optional static scan over multiple bed positions starting around 70 min. Kinetic analysis of lesion uptake was performed using blood-pool activity with population radiometabolite corrections. Normal breast and normal muscle reference tissue models were compared with PARP-1 protein expression in 10 patients with available tissue. Plasma radiometabolite concentrations and uptake in tumor and normal muscle were investigated in mouse xenografts. Results: Pharmacokinetics of ¹⁸F-FTT were well fit by Logan plot reference region models of reversible binding. However, fits of 2-tissue compartment models assuming negligible metabolite uptake were unstable. Rapid metabolism of ¹⁸F-FTT was demonstrated in mice, and similar uptake of radiometabolites was found in tumor xenografts and normal muscle. Tumor ¹⁸F-FTT distribution volume ratios relative to normal muscle reference tissue correlated with tissue PARP-1 expression (P < 0.02, n = 10). The tumor-to-normal muscle ratio from a 5-min frame between 50 and 60 min after injection, a potential static scan protocol, closely corresponded to the distribution volume ratio relative to normal muscle and correlated to PARP-1 expression (P < 0.02, n = 10). Conclusion: This study of PARPi analog ¹⁸F-FTT showed that uptake kinetics in vivo corresponded to expression of PARP-1 and that ¹⁸F-FTT quantitation is influenced by radiometabolites that are increasingly present late after injection. Radiometabolites can be controlled by using optimal image acquisition timing or normal muscle reference tissue modeling in dynamic imaging or a tumor-to-normal muscle ratio. Optimal image timing for tumor-to-normal muscle quantification in humans appears to be between 50 and 60 min after injection. Therefore, a clinically practical static imaging protocol commencing 45-55 min after injection may sufficiently balance ¹⁸F-FTT uptake with background clearance and radiometabolite interference for quantitative interpretation of PARP-1 expression in vivo.

Keywords: 18F-fluorthanatrace; PARP-1; breast cancer; radiotracer tissue pharmacokinetics.

Graphical abstract

FIGURE 1.

CT, PET, and fused images of ¹⁸F-FTT uptake in ER+ breast cancer in patient 6. Arrows indicate tumor on CT image (left) and corresponding PET and fused images (center and right). PET and fused images are scaled 0–5 g/mL, and CT image is scaled from −160 to +240 Hounsfield units.

FIGURE 2.

Individual tumor SUV_peak time–activity curves for 10 patients with surgical specimens.

FIGURE 3.

Tumor, blood-pool, and reference tissue time–activity curves for example patient (patient 9) with moderate ¹⁸F-FTT tumor uptake.

FIGURE 4.

Comparative imaging, staining, and radio–high-performance liquid chromatography (radioHPLC) analysis for HCC-1806 xenograft mouse. (A) ¹⁸F-fluorthanatrace PET/CT imaging. (B) Hematoxylin and eosin staining and immunohistochemical staining for PARP-1. (C) Tumor radioHPLC. (D) Parent fractions in sampled tissues and blood. Mouse was imaged 43–53 min after injection and killed for tissue sampling and radiometabolite analysis at 60 min. PET image is scaled 0–1 MBq/cm³, and CT image is scaled from −1,000 to +1,000 Hounsfield units. Tumor (arrow) had central necrosis, so periphery with elevated ¹⁸F-FTT uptake was used for radioHPLC.