Leveraging [¹⁸F]F-Piflufolastat PET for PBPK Prediction of Tissue Distribution and Therapy Optimization in Prostate...

Prostate-specific membrane antigen (PSMA) is overexpressed in over 90% of prostate cancer tumors, making it a strong target for imaging and therapy. The diagnostic tracer [¹⁸F]F-Piflufolastat provides high affinity and image quality, while the β-emitter [¹⁷⁷Lu]Lu-PSMA-617 enables targeted treatment. Despite differences in radioactive half-life, their shared PSMA-binding and pharmacokinetic profiles suggest potential for translational pharmacokinetic modeling. We hypothesized that the published full-body physiologically based pharmacokinetic (PBPK) model could be directly translated across the investigated PSMA ligands to predict tracer distribution within the same patient. A PBPK model was developed by rebuilding and modifying a published [⁶⁸Ga]Ga-PSMA-11/[¹⁷⁷Lu]Lu-PSMA-617 model to simulate [¹⁸F]F-Piflufolastat PK. Patient-specific factors included tumor volume, PSMA receptor density, organ blood flow, and renal clearance. Verification used published data, dynamic positron emission tomography (PET) from eight patients, and static total-body PET/computed tomography (CT) at 2 h from five patients. Extrapolation to [¹⁷⁷Lu]Lu-PSMA-617, using fitted receptor density, was evaluated against 24-h and 7-day Single-Photon Emission Computed Tomography (SPECT) data from two patients using mean prediction error (PE). The model accurately predicted [¹⁸F]F-Piflufolastat tumor uptake (median PE -3.44% (IQR -3.79 to -3.17) and -3.53% (IQR -3.78 to -3.00) in internal and external cohorts) and showed consistency with 7-day [¹⁷⁷Lu]Lu-PSMA-617 SPECT observations. However, predictions for other organs were variable (median PE ranging between -89.7 and 144), indicating limitations of a one-to-one molecular translational approach. These findings demonstrate that PBPK modeling can capture key determinants of PSMA tracer distribution, and link [¹⁸F]F-Piflufolastat imaging to [¹⁷⁷Lu]Lu-PSMA-617 predictions. However, improved representation of systemic and organ-level kinetics is required. Future work should explore a limited PBPK approach, focusing on accurate blood kinetics and organ-specific modeling.