Daan Van Valkengoed 1, Frida Bällgren 2,3, Vivi Rottschäfer 4,5, Irena Loryan 3, Elizabeth de Lange 1
1 Division of Systems Pharmacology and Pharmacy, Leiden Academic Center for Drug Research, Leiden University (Leiden, The Netherlands), 2 Department of Pharmacy, Science for Life Laboratory, Drug Discovery and Development (SciLifeLab DDD), Uppsala University (Uppsala, Sweden), 3 Translational Pharmacokinetics/Pharmacodynamics Group (tPKPD), Department of Pharmacy, Uppsala University (Uppsala, Sweden), 4 Mathematical Institute, Leiden University, Leiden (Leiden, The Netherlands), 5 Korteweg-de Vries Institute for Mathematics, University of Amsterdam (Amsterdam, The Netherlands)
Objectives: Poor understanding of central nervous system (CNS) target site pharmacokinetics (PK) is a major contributor to high attrition rates in CNS drug development (1). CNS drug exposure, especially of transporter substrates, is often unknown due to the CNS barriers modulating the rate and extent of drug disposition (2). This contributes to costly failures in late-phase drug discovery (1, 3). In vitro-in vivo extrapolation linked physiologically-based pharmacokinetic (IVIVE/PBPK) models hold great potential for advancing CNS drug discovery and development, as they allow for predictions of target-site PK early in the process based on in vitro data.
Predicting CNS PK requires a fundamental and integrated understanding of the blood-brain-barrier (BBB), blood-cerebrospinal fluid-barrier (BCSFB), IVIVE of transporter activity, and intra-brain distribution. The value and reliability of IVIVE/PBPK to predict CNS PK has been explored for passively diffusing and actively effluxed drugs (4–6). In recent years, the proton-coupled organic cation antiporter has been shown to be involved in active influx of many CNS-active drugs, including oxycodone and tramadol (7). Despite this, little is known about the translatability of in vitro antiporter activity into CNS PK in vivo through PBPK modelling. Given the ubiquity of the antiporter, a fundamental understanding of its activity and IVIVE is crucial to advance evidence-based CNS drug discovery and development. This study therefore aimed to evaluate the translatability of in vitro data on antiporter influx for PBPK predictions of oxycodone CNS disposition in rats.
Methods: This study used published microdialysis data of a 60-minute infusion of 0.3 mg/kg/h oxycodone in rats (8). Blood PK was modelled using Monolix (9), serving as input for the CNS PBPK model. The LeiCNS-PK3.4 PBPK model (6) was then applied for bottom-up prediction of oxycodone drug distribution in brain extracellular fluid (brainECF), lateral ventricles (LV) and cisterna magna (CM). Four literature-derived in vitro active uptake parameters of oxycodone quantified using Michaelis-Menten (MM) parameters (Jmax and Km) or apparent permeability (Papp) were used to predict oxycodone CNS disposition. The uptake parameters, all measured in rat or human-derived BBB cells, were scaled by the CNS barrier surface areas (SA) (6) or amount of protein in the BBB (10) depending on their units. The model’s accuracy in predicting oxycodone CNS PK was evaluated by the log2-ratios of the predicted AUC versus observed median AUC [log2(AUCpred/AUCobs)], predicted and observed concentrations at each time point [log2(Cpred/Cobs)], and median prediction errors. All metrics were deemed acceptable if within a 2-fold error. Lastly, the residual-mean-square-error (RMSE) was determined to facilitate comparison to the previously published LeiCNS-PK3.0 PBPK model which does not use in vitro data as input.
Results: LeiCNS-PK3.4 model predictions scaling transporter activity by surface area of the CNS barriers (using the Papp or one of the MM-parameter sets) showed a log2(AUCpred/AUCobs) of less than 2-fold (< 1 log2-unit) for all CNS locations (maximum deviation: brainECF, -0.41; LV, -0.37; CM, 0.86). Benchmarking these predictions to the LeiCNS-PK3.0 model showed similar RMSE values (e.g., LeiCNS-PK3.0, 35.6; SA-scaled MM LeiCNS-PK3.4, 36.3). Conversely, scaling the Jmax by BBB protein content (the remaining MM-parameter sets) gave poor predictions with maximum deviations of -1.61, 1.18, and 1.87 for brainECF, LV and CM. Strong shifts in the log2(Cpred/Cobs) over time for all predictions, especially those for the brainECF PK, indicate a mechanism of influx that is not yet accounted for. Comparing the predictive performance of the MM-parameters highlighted the superiority of the SA scaling to not be caused by a difference in the in vivo Jmax (SA-scaling, 5.4 nmol/min; protein-scaling, 6.1 nmol/min), but due to differences in Km (23 versus 89 µM). The intrinsic clearances (Jmax/Km) derived from these parameters as such vary greatly, e.g., 0.267 mL/min versus 0.06 mL/min. Conclusions: This CNS PBPK study is the first to evaluate the IVIVE/PBPK potential of antiporter-mediated drug influx at the CNS barriers. Scaling in vitro-derived active influx by barrier surface areas allowed for bottom-up predictions of oxycodone CNS PK at multiple CNS locations within two-fold error. The sensitivity of the model to the Km, rather than the scaled Jmax, underlines the impact of experimental uncertainty on IVIVE of MM-parameters. These results aid in fundamental understanding of transporter IVIVE, guiding the use of in vitro data to advance early CNS drug discovery and development. References: 1. Kesselheim AS, Hwang TJ, Franklin JM. 2015. Two decades of new drug development for central nervous system disorders. Nat Rev Drug Discov 14:815–816. 2. De Lange E, Hammarlund‐Udenaes M. 2015. Translational aspects of blood–brain barrier transport and central nervous system effects of drugs: From discovery to patients. Clin Pharmacol Ther 97:380–394. 3. Gribkoff VK, Kaczmarek LK. 2017. The need for new approaches in CNS drug discovery: Why drugs have failed, and what can be done to improve outcomes. Neuropharmacology 120:11–19. 4. Uchida Y, Ohtsuki S, Kamiie J, Terasaki T. 2011. Blood-Brain Barrier (BBB) Pharmacoproteomics: Reconstruction of In Vivo Brain Distribution of 11 P-Glycoprotein Substrates Based on the BBB Transporter Protein Concentration, In Vitro Intrinsic Transport Activity, and Unbound Fraction in Plasma and Brain in Mice. J Pharmacol Exp Ther 339:579–588. 5. Storelli F, Anoshchenko O, Unadkat JD. 2021. Successful Prediction of Human Steady‐State Unbound Brain‐to‐Plasma Concentration Ratio of P‐gp Substrates Using the Proteomics‐Informed Relative Expression Factor Approach. Clin Pharmacol Ther 110:432–442. 6. Van Valkengoed DW, Hirasawa M, Rottschäfer V, De Lange ECM. 2025. Reliability of in vitro data for the mechanistic prediction of brain extracellular fluid pharmacokinetics of P-glycoprotein substrates in vivo; are we scaling correctly? J Pharmacokinet Pharmacodyn 52:16. 7. Sachkova A, Jensen O, Dücker C, Ansari S, Brockmöller J. 2022. The mystery of the human proton-organic cation antiporter: One transport protein or many? Pharmacol Ther 239:108283. 8. Bällgren F. 2024. Translational Aspects of Brain-Specific Drug Delivery by Targeting Active Uptake at Brain Barriers. Acta Universitatis Upsaliensis, Uppsala. 9. Simulations Plus (United States). 2024. Monolix Documentation (2024R1). 10. Verscheijden LFM, Litjens CHC, Koenderink JB, Mathijssen RHJ, Verbeek MM, De Wildt SN, Russel FGM. 2021. Physiologically based pharmacokinetic/pharmacodynamic model for the prediction of morphine brain disposition and analgesia in adults and children. PLOS Comput Biol 17:e1008786.
Reference: PAGE 34 (2026) Abstr 11945 [www.page-meeting.org/?abstract=11945]
Poster: Drug/Disease Modelling - Absorption & PBPK