Associate Scientist Certara Rolesville, North Carolina, United States
Disclosure(s):
Hunter Stephens, PhD: No relevant disclosure to display
Objectives: While targeted radiation therapies (TRTs) are raising more and more interest in the oncology space, the regulatory landscape continues to evolve leaving development teams without clear guidance. Model-Informed Drug Development (MIDD) can provide key insights into optimizing doses to achieve maximum efficacy within safety bounds, as well as the design of novel therapeutics with enhanced therapeutic index. The theranostic nature of TRTs allows for precise modelling and prediction of drug kinetics and absorbed doses which will allow for better decision-making in the development process. In this abstract, we present an end-to-end framework for using MIDD with TRTs, specifically with PSMA targeting radio-ligands.
Methods: Taking both plasma activity samples and combining with nuclear imaging, the drug kinetics was modelled in both the blood and organs of interest. The semi-mechanistic compartmental models incorporated any saturable uptake of the radio-ligand into peripheral organs of interest as a function of tissue volume and target expression. Parameterizing the models in terms of drug radioactivity allowed for the calculation and prediction of absorbed doses into specific organs of interest. These models were used to optimize drug properties such as binding affinity, and plasma half-life to improve therapeutic index and support candidate selection. Subsequently, at later stages, with the ability to predict organ absorbed dose, the dosing regimen can be optimized by minimizing the simulated adverse event profile linked to specific organs while maximizing tumour exposure.
Results: The MIDD framework was validated using published clinical data with published Pluvicto (lutetium Lu 177 vipivotide tetraxetan). The framework made accurate predictions of across a range of doses, PSMA densities, and tumor volumes accounting for distribution and competitive binding of labeled and un-labeled drug in tumor and non-tumor tissues. In addition, the framework allowed to characterize the specific radioactive dose absorbed into blood and various organs such as liver, kidney, spleen, salivary gland and tumor. The absorbed dose predictions were in good alignment with the patient specific dosimetry results. Subsequently, the model predicted exposures were linked to safety and efficacy endpoints which allowed extrapolation from the dosimetry to clinical endpoints.
Conclusions: The MIDD framework can be utilized to translate drug design features such as receptor binding affinity on the disposition in various organs of the radio-labeled drug to inform clinical candidate selection to help maximize the therapeutic index. Subsequently, the framework can be utilized to establish first-in-human dosing regimens and to optimize recommended Phase 2 doses.
