Vivi Rottschäfer
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Vivi Rottschäfer has a total of 51 co-authors. Here are the top co-authors by number of shared publications:
- Elizabeth Cunera Maria de Lange
Division of Systems Pharmacology and Pharmacy, Leiden Academic Centre for Drug Research Leiden University Leiden The Netherlands
12 shared publications - Esmée Vendel
Affiliation Mathematical Institute, Leiden University, Leiden, The Netherlands
5 shared publications - Mengxu Zhang
Division of Systems Pharmacology and Pharmacy, Predictive Pharmacology group, Leiden Academic Centre of Drug Research, Leiden University, Leiden, The Netherlands
3 shared publications - Carla White
Swansea University, Swansea, UK
2 shared publications - Daan W. van Valkengoed
Division of Systems Pharmacology and Pharmacy, Leiden Academic Centre for Drug Research, Leiden University, Leiden, The Netherlands
2 shared publications - Ed VanBavel
Department of Biomedical Engineering and Physics, Amsterdam UMC, University of Amsterdam, Amsterdam, Netherlands
2 shared publications
Latest Publications
- Open Access
- 23 October 2026
- Published by Elsevier in European Journal of Pharmaceutical Sciences
- Vol. 228, 107696
Application of the LeiCNS-PK model to quantitatively evaluate the roles of BBB active influx and brain metabolism of fluvoxamine and its serotonin transporter (SERT) occupancy in rat and human
- 23 October 2026
- Published by Elsevier in European Journal of Pharmaceutical Sciences
- Vol. 228, 107696
Open AccessApplication of the LeiCNS-PK model to quantitatively evaluate the roles of BBB active influx and brain metabolism of fluvoxamine and its serotonin transporter (SERT) occupancy in rat and human
- Open Access
- 9 September 2026
- Published by Wiley in British Journal of Clinical Pharmacology
The integrated Lei‐CNS PBPK‐PD model effectively predicts clozapine brain disposition and D 2 receptor occupancy in humans
Aim: Clozapine, used for treatment‐resistant schizophrenia, has a narrow therapeutic window and highly variable plasma pharmacokinetics (PK), necessitating careful monitoring. Direct measurement of unbound brain PK in humans is unethical. To develop a physiologically based pharmacokinetic–pharmacodynamic (PBPK‐PD) model to predict clozapine brain extracellular fluid (brainECF) PK in rats and humans and D 2 receptor occupancy (RO) in humans. Methods: A previously developed rat CNS PBPK model (LeiCNS‐PK3.5), containing brain metabolism, was extended to include active blood–brain barrier (BBB) influx transport and neutral lipid binding (LeiCNS‐PK3.6). LeiCNS‐PK3.6 was validated against observed brainECF clozapine concentrations in rats. For human translation, species‐specific transporter and enzyme parameters were applied. A receptor binding kinetics model, with association and dissociation rate constants, was developed to predict D 2 RO based on unbound brainECF clozapine concentrations and validated using PET‐derived clinical data. Results: The LeiCNS‐PK3.6 model significantly improved the prediction accuracy of brainECF clozapine PK in rats compared with previous versions. In humans, predicted brainECF concentrations aligned with expected therapeutic ranges. The PBPK‐PD model accurately predicted D 2 RO in most patients, suggesting polymorphisms, smoking and dissociation rate constant (koff) may affect prediction accuracy. Conclusions: The integrated CNS PBPK‐PD model effectively predicts clozapine brain disposition and D 2 RO in humans. This framework offers a valuable tool to explore inter‐individual variability (e.g., CYP1A2 polymorphisms, smoking and koff), supporting personalized dosing strategies and improved safety in clozapine therapy.
- 9 September 2026
- Published by Wiley in British Journal of Clinical Pharmacology
Open AccessThe integrated Lei‐CNS PBPK‐PD model effectively predicts clozapine brain disposition and D 2 receptor occupancy in humans
Aim: Clozapine, used for treatment‐resistant schizophrenia, has a narrow therapeutic window and highly variable plasma pharmacokinetics (PK), necessitating careful monitoring. Direct measurement of unbound brain PK in humans is unethical. To develop a physiologically based pharmacokinetic–pharmacodynamic (PBPK‐PD) model to predict clozapine brain extracellular fluid (brainECF) PK in rats and humans and D 2 receptor occupancy (RO) in humans. Methods: A previously developed rat CNS PBPK model (LeiCNS‐PK3.5), containing brain metabolism, was extended to include active blood–brain barrier (BBB) influx transport and neutral lipid binding (LeiCNS‐PK3.6). LeiCNS‐PK3.6 was validated against observed brainECF clozapine concentrations in rats. For human translation, species‐specific transporter and enzyme parameters were applied. A receptor binding kinetics model, with association and dissociation rate constants, was developed to predict D 2 RO based on unbound brainECF clozapine concentrations and validated using PET‐derived clinical data. Results: The LeiCNS‐PK3.6 model significantly improved the prediction accuracy of brainECF clozapine PK in rats compared with previous versions. In humans, predicted brainECF concentrations aligned with expected therapeutic ranges. The PBPK‐PD model accurately predicted D 2 RO in most patients, suggesting polymorphisms, smoking and dissociation rate constant (koff) may affect prediction accuracy. Conclusions: The integrated CNS PBPK‐PD model effectively predicts clozapine brain disposition and D 2 RO in humans. This framework offers a valuable tool to explore inter‐individual variability (e.g., CYP1A2 polymorphisms, smoking and koff), supporting personalized dosing strategies and improved safety in clozapine therapy.
- Open Access
- 2 February 2026
- Published by Springer Nature in Journal of Pharmacokinetics and Pharmacodynamics
- Vol. 53 (2) , 10
Simulation-based assessment of the P-glycoprotein expression-activity relationship shows a drug and system dependency
In vitro-in vivo extrapolation (IVIVE) of P-glycoprotein (P-gp) transporter activity bears two important assumptions: (1) P-gp expression (i.e., concentration) is linearly related to P-gp activity; and (2) this relation is drug independent. However, conflicting experimental results have been obtained about this relationship. This study therefore aimed to theoretically explore the P-gp expression-activity relationship (EAR). A P-gp membrane kinetic binding model was used to explore the P-gp EAR, and how this is impacted by different drugs as well as drug dose and initial P-gp expression. The model included passive permeability and drug interaction with P-gp described through the association ( k on ), dissociation ( k off ), and efflux rate ( k e ) constants. Simulations were first performed for 7 P-gp substrates, assuming an initial P-gp expression of 1000 µM and drug concentrations of 1 µM. Subsequently, the P-gp concentration was varied (2-300%) to derive the EAR. Next, to determine the impact of drug-specific and system-dependent properties on the P-gp EAR, virtual drugs varying in k off and k e were simulated, simultaneously considering different initial P-gp expressions and drug concentrations. Our simulations show that the P-gp EAR is not always linear, with the EAR showing both linear and non-linear behaviour depending on the drug. The k off /k e ratio of a drug was found to be an important determinant of the P-gp EAR, which shifted towards non-linearity for lower k off / k e ratios. Additionally, the P-gp EAR was more likely to be non-linear for higher ratios of the initial P-gp expression to drug concentration. In conclusion, this study shows that P-gp expression is not always proportional to P-gp activity, thereby explaining experimental contradictions about the EAR. This implies that using P-gp expression as single biomarker for P-gp activity needs to be reconsidered.
- 2 February 2026
- Published by Springer Nature in Journal of Pharmacokinetics and Pharmacodynamics
- Vol. 53 (2) , 10
Open AccessSimulation-based assessment of the P-glycoprotein expression-activity relationship shows a drug and system dependency
In vitro-in vivo extrapolation (IVIVE) of P-glycoprotein (P-gp) transporter activity bears two important assumptions: (1) P-gp expression (i.e., concentration) is linearly related to P-gp activity; and (2) this relation is drug independent. However, conflicting experimental results have been obtained about this relationship. This study therefore aimed to theoretically explore the P-gp expression-activity relationship (EAR). A P-gp membrane kinetic binding model was used to explore the P-gp EAR, and how this is impacted by different drugs as well as drug dose and initial P-gp expression. The model included passive permeability and drug interaction with P-gp described through the association ( k on ), dissociation ( k off ), and efflux rate ( k e ) constants. Simulations were first performed for 7 P-gp substrates, assuming an initial P-gp expression of 1000 µM and drug concentrations of 1 µM. Subsequently, the P-gp concentration was varied (2-300%) to derive the EAR. Next, to determine the impact of drug-specific and system-dependent properties on the P-gp EAR, virtual drugs varying in k off and k e were simulated, simultaneously considering different initial P-gp expressions and drug concentrations. Our simulations show that the P-gp EAR is not always linear, with the EAR showing both linear and non-linear behaviour depending on the drug. The k off /k e ratio of a drug was found to be an important determinant of the P-gp EAR, which shifted towards non-linearity for lower k off / k e ratios. Additionally, the P-gp EAR was more likely to be non-linear for higher ratios of the initial P-gp expression to drug concentration. In conclusion, this study shows that P-gp expression is not always proportional to P-gp activity, thereby explaining experimental contradictions about the EAR. This implies that using P-gp expression as single biomarker for P-gp activity needs to be reconsidered.
- 10 April 2025
- Published by American Physical Society (APS) in Physical Review E
- Vol. 111 (4) , 044406
Messenger size optimality in cellular communications
Living cells presumably employ optimized information transfer methods, enabling efficient communication even in noisy environments. The efficiency of chemical communication between cells depends on the properties of the molecular messenger. Evidence suggests that proteins within narrow molecular mass ranges have been naturally selected to mediate cellular communication, yet the underlying communication design principles remain poorly understood. Using a simple physical model that accounts for the cost of chemical synthesis, diffusion, molecular binding, and degradation, we show that optimal masses exist to ensure efficient communication of various types of signals. Our findings provide insights into the design principles of biological communication and may inform the engineering of biomimetic systems with chemical signaling capabilities.
- 10 April 2025
- Published by American Physical Society (APS) in Physical Review E
- Vol. 111 (4) , 044406
Messenger size optimality in cellular communications
Living cells presumably employ optimized information transfer methods, enabling efficient communication even in noisy environments. The efficiency of chemical communication between cells depends on the properties of the molecular messenger. Evidence suggests that proteins within narrow molecular mass ranges have been naturally selected to mediate cellular communication, yet the underlying communication design principles remain poorly understood. Using a simple physical model that accounts for the cost of chemical synthesis, diffusion, molecular binding, and degradation, we show that optimal masses exist to ensure efficient communication of various types of signals. Our findings provide insights into the design principles of biological communication and may inform the engineering of biomimetic systems with chemical signaling capabilities.
- Open Access
- 8 April 2025
- Published by Elsevier in Journal of Theoretical Biology
- Vol. 607, 112111
Arterial arcades and collaterals regress under hemodynamics-based diameter adaptation: A computational and mathematical analysis
- 8 April 2025
- Published by Elsevier in Journal of Theoretical Biology
- Vol. 607, 112111
Open AccessArterial arcades and collaterals regress under hemodynamics-based diameter adaptation: A computational and mathematical analysis
- Open Access
- 8 February 2025
- Published by Springer Nature in Journal of Pharmacokinetics and Pharmacodynamics
- Vol. 52 (2) , 1-20
Reliability of in vitro data for the mechanistic prediction of brain extracellular fluid pharmacokinetics of P-glycoprotein substrates in vivo; are we scaling correctly?
Plasma pharmacokinetic (PK) profiles often do not resemble the PK within the central nervous system (CNS) because of blood–brain-border (BBB) processes, like active efflux by P-glycoprotein (P-gp). Methods to predict CNS-PK are therefore desired. Here we investigate whether in vitro apparent permeability (Papp) and corrected efflux ratio (ERc) extracted from literature can be repurposed as input for the LeiCNS-PK3.4 physiologically-based PK model to confidently predict rat brain extracellular fluid (ECF) PK of P-gp substrates. Literature values of in vitro Caco-2, LLC-PK1-mdr1a/MDR1, and MDCKII-MDR1 cell line transport data were used to calculate P-gp efflux clearance (CLPgp). Subsequently, CLPgp was scaled from in vitro to in vivo through a relative expression factor (REF) based on P-gp expression differences. BrainECF PK was predicted well (within twofold error of the observed data) for 2 out of 4 P-gp substrates after short infusions and 3 out of 4 P-gp substrates after continuous infusions. Variability of in vitro parameters impacted both predicted rate and extent of drug distribution, reducing model applicability. Notably, use of transport data and in vitro P-gp expression obtained from a single study did not guarantee an accurate prediction; it often resulted in worse predictions than when using in vitro expression values reported by other labs. Overall, LeiCNS-PK3.4 shows promise in predicting brainECF PK, but this study highlights that the in vitro to in vivo translation is not yet robust. We conclude that more information is needed about context and drug dependency of in vitro data for robust brainECF PK predictions.
- 8 February 2025
- Published by Springer Nature in Journal of Pharmacokinetics and Pharmacodynamics
- Vol. 52 (2) , 1-20
Open AccessReliability of in vitro data for the mechanistic prediction of brain extracellular fluid pharmacokinetics of P-glycoprotein substrates in vivo; are we scaling correctly?
Plasma pharmacokinetic (PK) profiles often do not resemble the PK within the central nervous system (CNS) because of blood–brain-border (BBB) processes, like active efflux by P-glycoprotein (P-gp). Methods to predict CNS-PK are therefore desired. Here we investigate whether in vitro apparent permeability (Papp) and corrected efflux ratio (ERc) extracted from literature can be repurposed as input for the LeiCNS-PK3.4 physiologically-based PK model to confidently predict rat brain extracellular fluid (ECF) PK of P-gp substrates. Literature values of in vitro Caco-2, LLC-PK1-mdr1a/MDR1, and MDCKII-MDR1 cell line transport data were used to calculate P-gp efflux clearance (CLPgp). Subsequently, CLPgp was scaled from in vitro to in vivo through a relative expression factor (REF) based on P-gp expression differences. BrainECF PK was predicted well (within twofold error of the observed data) for 2 out of 4 P-gp substrates after short infusions and 3 out of 4 P-gp substrates after continuous infusions. Variability of in vitro parameters impacted both predicted rate and extent of drug distribution, reducing model applicability. Notably, use of transport data and in vitro P-gp expression obtained from a single study did not guarantee an accurate prediction; it often resulted in worse predictions than when using in vitro expression values reported by other labs. Overall, LeiCNS-PK3.4 shows promise in predicting brainECF PK, but this study highlights that the in vitro to in vivo translation is not yet robust. We conclude that more information is needed about context and drug dependency of in vitro data for robust brainECF PK predictions.
- Open Access
- 11 September 2024
- Published by Society for Neuroscience in eneuro
- Vol. 11 (9)
The Neural and Computational Architecture of Feedback Dynamics in Mouse Cortex during Stimulus Report
Conscious reportability of visual input is associated with a bimodal neural response in the primary visual cortex (V1): an early-latency response coupled to stimulus features and a late-latency response coupled to stimulus report or detection. This late wave of activity, central to major theories of consciousness, is thought to be driven by the prefrontal cortex (PFC), responsible for “igniting” it. Here we analyzed two electrophysiological studies in mice performing different stimulus detection tasks and characterized neural activity profiles in three key cortical regions: V1, posterior parietal cortex (PPC), and PFC. We then developed a minimal network model, constrained by known connectivity between these regions, reproducing the spatiotemporal propagation of visual- and report-related activity. Remarkably, while PFC was indeed necessary to generate report-related activity in V1, this occurred only through the mediation of PPC. PPC, and not PFC, had the final veto in enabling the report-related late wave of V1 activity.
- 11 September 2024
- Published by Society for Neuroscience in eneuro
- Vol. 11 (9)
Open AccessThe Neural and Computational Architecture of Feedback Dynamics in Mouse Cortex during Stimulus Report
Conscious reportability of visual input is associated with a bimodal neural response in the primary visual cortex (V1): an early-latency response coupled to stimulus features and a late-latency response coupled to stimulus report or detection. This late wave of activity, central to major theories of consciousness, is thought to be driven by the prefrontal cortex (PFC), responsible for “igniting” it. Here we analyzed two electrophysiological studies in mice performing different stimulus detection tasks and characterized neural activity profiles in three key cortical regions: V1, posterior parietal cortex (PPC), and PFC. We then developed a minimal network model, constrained by known connectivity between these regions, reproducing the spatiotemporal propagation of visual- and report-related activity. Remarkably, while PFC was indeed necessary to generate report-related activity in V1, this occurred only through the mediation of PPC. PPC, and not PFC, had the final veto in enabling the report-related late wave of V1 activity.
- Open Access
Arterial arcades and collaterals regress under hemodynamics-based diameter adaptation: a computational and mathematical analysis
Segments in the arterial network have a >1000-fold span of radii. This is believed to result from adaptation of each segment to the wall shear stress (WSS), with outward respectively inward remodeling if WSS is higher or lower than some reference value. While this seems a straightforward mechanism for arterial tree design, the arterial network is not a tree but contains numerous arcades, collaterals and other looping structures. In this theoretical study, we analyzed stability of looping structures in arterial networks under WSS control. Simulation models were based on very simple network topologies as well as on published human coronary and mouse cerebral arterial networks. Adaptation was implemented as a rate of change of structural radius of each segment that is proportional to the deviation from its reference WSS. A more generalized model was based on adaptation to a large range of other local hemodynamic stimuli, including velocity, flow and power dissipation. For over 12,000 tested parameter sets, the simulations invariably predicted loss of loops due to regression of one or more of the segments. In the small networks, this was the case for both the WSS and the generalized model, and for a large range of initial conditions and model parameters. Loss of loopiness also was predicted by models that included direction-dependent adaptation rates, heterogeneous reference WSS or adaptation rates among the adapting segments, and adaptation under dynamic conditions. Loss of loops was also found in the coronary and cerebral artery networks subjected to adaptation to WSS. In a mathematical analysis we proved that loss of loops is a direct consequence of Kirchhoff’s circuit law, which for each loop leads to a positive eigenvalue in the Jacobian matrix of partial derivatives in the adaptation model, and therefore to unstable equilibria in the presence of loops. Loss of loops is an inherent property of arterial networks that adapt to local hemodynamics. Additional mechanisms are therefore needed to explain their presence, including direct communication between connected segments.
Open AccessArterial arcades and collaterals regress under hemodynamics-based diameter adaptation: a computational and mathematical analysis
Segments in the arterial network have a >1000-fold span of radii. This is believed to result from adaptation of each segment to the wall shear stress (WSS), with outward respectively inward remodeling if WSS is higher or lower than some reference value. While this seems a straightforward mechanism for arterial tree design, the arterial network is not a tree but contains numerous arcades, collaterals and other looping structures. In this theoretical study, we analyzed stability of looping structures in arterial networks under WSS control. Simulation models were based on very simple network topologies as well as on published human coronary and mouse cerebral arterial networks. Adaptation was implemented as a rate of change of structural radius of each segment that is proportional to the deviation from its reference WSS. A more generalized model was based on adaptation to a large range of other local hemodynamic stimuli, including velocity, flow and power dissipation. For over 12,000 tested parameter sets, the simulations invariably predicted loss of loops due to regression of one or more of the segments. In the small networks, this was the case for both the WSS and the generalized model, and for a large range of initial conditions and model parameters. Loss of loopiness also was predicted by models that included direction-dependent adaptation rates, heterogeneous reference WSS or adaptation rates among the adapting segments, and adaptation under dynamic conditions. Loss of loops was also found in the coronary and cerebral artery networks subjected to adaptation to WSS. In a mathematical analysis we proved that loss of loops is a direct consequence of Kirchhoff’s circuit law, which for each loop leads to a positive eigenvalue in the Jacobian matrix of partial derivatives in the adaptation model, and therefore to unstable equilibria in the presence of loops. Loss of loops is an inherent property of arterial networks that adapt to local hemodynamics. Additional mechanisms are therefore needed to explain their presence, including direct communication between connected segments.
- Open Access
Exploring Kp,uu,BBB Values Smaller than Unity in Remoxipride: A Physiologically-Based CNS Model Approach Highlighting Brain Metabolism in Drugs with Passive Blood-Brain Barrier Transport
(Aim): K p,uu,BBB values are crucial indicators of drug distribution into the brain, representing the steady-state relationship between unbound concentrations in plasma and in brain extracellular fluid (brainECF). K p,uu,BBB values < 1 are often interpreted as indicators of dominant active efflux transport processes at the blood-brain barrier (BBB). However, the potential impact of brain metabolism on this value is typically not addressed. In this study, we investigated the brain distribution of remoxipride, as a paradigm compound for passive BBB transport with yet unexplained brain elimination that was hypothesized to represent brain metabolism. (Methods): The physiologically-based LeiCNS pharmacokinetic predictor (LeiCNS-PK model) was used to compare brain distribution of remoxipride with and without Michaelis-Menten kinetics at the BBB and/or brain cell organelle levels. To that end, multiple in-house (IV 0.7, 3.5, 4, 5.2, 7, 8, 14 and 16 mg/kg) and external (IV 4 and 8 mg/kg) rat microdialysis studies plasma and brainECF data were analysed. (Results): The incorporation of active elimination through presumed brain metabolism of remoxipride in the LeiCNS-PK model significantly improved the prediction accuracy of experimentally observed brainECF profiles of this drug. The model integrated with brain metabolism in both barriers and organelles levels is named LeiCNS-PK3.5. (Conclusion): For drugs with K p,uu,BBB values < 1, not only the current interpretation of dominant BBB efflux transport, but also potential brain metabolism needs to be considered, especially because these may be concentration dependent. This will improve the mechanistic understanding of the processes that determine brain PK profiles.
Open AccessExploring Kp,uu,BBB Values Smaller than Unity in Remoxipride: A Physiologically-Based CNS Model Approach Highlighting Brain Metabolism in Drugs with Passive Blood-Brain Barrier Transport
(Aim): K p,uu,BBB values are crucial indicators of drug distribution into the brain, representing the steady-state relationship between unbound concentrations in plasma and in brain extracellular fluid (brainECF). K p,uu,BBB values < 1 are often interpreted as indicators of dominant active efflux transport processes at the blood-brain barrier (BBB). However, the potential impact of brain metabolism on this value is typically not addressed. In this study, we investigated the brain distribution of remoxipride, as a paradigm compound for passive BBB transport with yet unexplained brain elimination that was hypothesized to represent brain metabolism. (Methods): The physiologically-based LeiCNS pharmacokinetic predictor (LeiCNS-PK model) was used to compare brain distribution of remoxipride with and without Michaelis-Menten kinetics at the BBB and/or brain cell organelle levels. To that end, multiple in-house (IV 0.7, 3.5, 4, 5.2, 7, 8, 14 and 16 mg/kg) and external (IV 4 and 8 mg/kg) rat microdialysis studies plasma and brainECF data were analysed. (Results): The incorporation of active elimination through presumed brain metabolism of remoxipride in the LeiCNS-PK model significantly improved the prediction accuracy of experimentally observed brainECF profiles of this drug. The model integrated with brain metabolism in both barriers and organelles levels is named LeiCNS-PK3.5. (Conclusion): For drugs with K p,uu,BBB values < 1, not only the current interpretation of dominant BBB efflux transport, but also potential brain metabolism needs to be considered, especially because these may be concentration dependent. This will improve the mechanistic understanding of the processes that determine brain PK profiles.
- Open Access
- 17 May 2024
- Published by Cambridge University Press (CUP) in Mathematics in Industry Repository
Frequency Domain Characteristics in Nonlinear Trimmer Dynamics
We analyse a dynamical system describing the movement of the blade of a hair trimmer. Using nondimensionalisation and perturbation methods, we systematically analyse this model, with the goal of explaining a dip in the current amplitude observed in numerical simulations. We find that the motion of the blade is controlled by seven dimensionless parameters, and the dip in amplitude of the current corresponds to one of the dimensionless parameter groups taking the value of one. There is excellent agreement between our asymptotic solutions and long-time behaviour of the model revealed by numerical simulation.
- 17 May 2024
- Published by Cambridge University Press (CUP) in Mathematics in Industry Repository
Open AccessFrequency Domain Characteristics in Nonlinear Trimmer Dynamics
We analyse a dynamical system describing the movement of the blade of a hair trimmer. Using nondimensionalisation and perturbation methods, we systematically analyse this model, with the goal of explaining a dip in the current amplitude observed in numerical simulations. We find that the motion of the blade is controlled by seven dimensionless parameters, and the dip in amplitude of the current corresponds to one of the dimensionless parameter groups taking the value of one. There is excellent agreement between our asymptotic solutions and long-time behaviour of the model revealed by numerical simulation.