Nuventra is proud to announce that it was awarded a grant from the U.S. Food and Drug Administration (FDA) for the “Development of machine learning approaches to population pharmacokinetic model selection and evaluation of application to model-based bioequivalence analysis.” The grant has been awarded in collaboration with the School of Pharmacy and Pharmaceutical Sciences of the State University of New York at Buffalo.
The purpose of the project is to use deep-learning / reinforcement-learning based algorithms to develop a method for selecting the most effective model for conducting population pharmacokinetic (popPK) analyses of bioequivalence (BE) studies. Selecting an appropriate model is key to a successful popPK analysis and popPK analyses are crucial in the emerging technology of model-based BE analysis.
PopPK modeling helps us understand the variability in drug concentrations among individuals in a group of interest (the “population”). Using modeling techniques such as popPK analysis allows drug developers to create safer and more effective dosing regimens, ultimately leading to safer and more effective therapies for clinical trial participants and patients.
Model-based BE analyses are essential for hard-to-study patient populations such as those with rare diseases or those utilizing drugs or drug device combinations with complex and long-acting delivery where traditional BE clinical studies are not feasible. For generic, investigational drugs, BE studies allow drug developers to confirm that the rate and extent of drug absorption does not differ significantly from the reference listed drug (RLD).
Nuventra’s Mark Sale, M.D., Executive Vice President of Pharmacometrics and University of Buffalo’s Robert Bies Pharm.D., Ph.D., Professor of the Department of Pharmaceutical Sciences will both serve as investigators on this project.
“Novel approaches to demonstrating BE have been used for several drug products in recent years, particularly characterization-based BE approaches, which are very different compared to traditional comparative PK or PD study approaches. The traditional approach presents challenges for a lot of drugs either with very long half-lives, or when samples are difficult to obtain,” said Dr. Sale, the principal investigator, further adding, “the FDA is interested in applying new methods to address these issues, and Nuventra is looking forward to helping develop model-based approaches to do that.”
Dr. Bies added that “The machine learning approaches (genetic algorithm [GA] and deep Q-learning [DQN]) that systematize the development of models from data for the determination of drug exposure has the potential to significantly reduce bias in generating metrics for exposure used for bioequivalence.”
*The work is funded 100% with a grant from the Federal Government, in the amount of $125,000
Nuventra is now part of CATO SMS, a global provider of clinical research solutions including regulatory consulting and full-service clinical trial operations with a focus on small and emerging companies. The acquisition is part of CATO SMS’ strategic expansion to better fulfill the drug development needs of biopharmaceutical companies from strategy to approval. The acquisition of Nuventra extends CATO SMS’ offering into the critical area of clinical pharmacology and modeling and simulation. Learn more about CATO SMS’ acquisition of Nuventra.