Anu is a fourth-year PhD candidate at the University of California San Francisco under the mentorship of Dr. Rada Savic. Prior to her graduate work, she completed a PharmD at the University of Pittsburgh. Currently, she combines her extensive pharmacology background and passion for improving patient outcomes with her computational modeling abilities to help create translational tools to accelerate anti-tuberculosis drug development.
Impact details available upon request.
Source: Conference 2024Having a standardized approach for human dose prediction (HDP) early in discovery and development is transformative for increasing the likelihood of a compound’s success. In the tuberculosis (TB) field, plasma exposure over the minimum inhibitory concentration (MIC) has been used as a pharmacokinetic-pharmacodynamic (PK-PD) target to inform efficacious dose, but this has not aligned with clinical results. By performing a back-translational PK-PD analysis of approved anti-TB drugs, we aim to identify best methods for HDP.Methods: Potency assay data, as well as translational and clinical PK models, were collected for ten anti-TB drugs, including four that underwent clinical dose-finding studies. Exposures in various matrices (plasma, lung, lesions, cavitary caseum) based on a rabbit-to-human translational tool were simulated across doses up to 10000 mg using NONMEM. Coverage over relevant PD targets, including MIC, intramacrophage IC90, caseum MBC90, and in vivo EC50, was simulated and an effective dose quantified, defined as achieving coverage over 50% of the steady-state dosing interval (ED50). This ED50 was then compared to the currently known efficacious dose of each drug.Results: Generally, plasma PK over MIC underpredicted the clinical efficacious dose across drugs, even after protein binding corrections. Lung PK over EC50, a dynamic metric quantified from in vivo PK-PD studies, resulted in predictions within approximately two-fold of efficacious dose for clinically dose-optimized drugs. Assessing caseum PK-PD is essential for identifying efficacy in hard-to-treat patient phenotypes and could accurately identify penetrating drugs.Conclusions: Efficacious dose prediction relies on the attainment of an accurate PK-PD target. Coverage defined by site-of-disease exposures over a dynamic potencytarget has been shown to be most predictive of clinical efficacious dose. This approach has the potential to reduce the time needed for a drug to reach clinical stages with an acceptable dose range and minimize costly dose-finding studies.
Source: Conference 2024