The mastermind approach to CNS drug therapy: translational prediction of human brain distribution, target site kinetics, and therapeutic effects

Abstract

Despite enormous advances in CNS research, CNS disorders remain the world's leading cause of disability. This accounts for more hospitalizations and prolonged care than almost all other diseases combined, and indicates a high unmet need for good CNS drugs and drug therapies.Following dosing, not only the chemical properties of the drug and blood-brain barrier (BBB) transport, but also many other processes will ultimately determine brain target site kinetics and consequently the CNS effects. The rate and extent of all these processes are regulated dynamically, and thus condition dependent. Therefore, heterogenious conditions such as species, gender, genetic background, tissue, age, diet, disease, drug treatment etc., result in considerable inter-individual and intra-individual variation, often encountered in CNS drug therapy.For effective therapy, drugs should access the CNS "at the right place, at the right time, and at the right concentration". To improve CNS therapies and drug development, details of inter-species and inter-condition variations are needed to enable target site pharmacokinetics and associated CNS effects to be translated between species and between disease states. Specifically, such studies need to include information about unbound drug concentrations which drive the effects. To date the only technique that can obtain unbound drug concentrations in brain is microdialysis. This (minimally) invasive technique cannot be readily applied to humans, and we need to rely on translational approaches to predict human brain distribution, target site kinetics, and therapeutic effects of CNS drugs.In this review the term "Mastermind approach" is introduced, for strategic and systematic CNS drug research using advanced preclinical experimental designs and mathematical modeling. In this way, knowledge can be obtained about the contributions and variability of individual processes on the causal path between drug dosing and CNS effect in animals that can be translated to the human situation. On the basis of a few advanced preclinical microdialysis based investigations it will be shown that the "Mastermind approach" has a high potential for the prediction of human CNS drug effects.

Figure 1

Schematic presentation of the major compartments of the mammalian brain and routes for drug exchange; extracellular fluid (ECF), brain cells, lateral ventricular CSF, cisterna magna CSF and lumbar CSF, passive transport (black arrows) and active transport (white arrows), as well as metabolism and CSF turnover. Drug targets may be present at different sites within the brain.

Figure 2

Factors affecting the pharmacokinetics and pharmacodynamics of a drug. The effects of a drug are determined on one hand by its physico-chemical/molecular characteristics and on the other hand by the properties of the biological systems involved.

Figure 3

Longitudinal section of the rat brain (From: Paxinos and Watson [136]) with the positions of the microdialysis probes indicated. From left to right: probe position in striatum, lateral ventricle CSF and cisterna magna CSF, respectively.

Figure 4

Brain distribution of acetaminophen in the rat. a) Data obtained for acetaminophen in the rat following an intravenous dose of 15 mg/kg, administered by constant-rate infusion for 10 minutes. The data are presented as the average (geometric mean ± SEM) of the observed unbound acetaminophen concentration-time profiles in plasma (black, n = 10), striatum ECF (ST, blue, n = 10), CSF in lateral ventricle (LV, green, n = 14), and CSF in cisterna magna (CM, red, n = 8). The data show that brain ECF (striatum) concentrations are comparable to those in plasma and significantly higher than those in both the lateral ventricle and the cisterna magna CSF compartments. b) The physiologically-based pharmacokinetic model for the rat developed on the basis of the data obtained for acetaminophen as shown in a). This model describes the obtained data adequately, and predicts the CSF acetaminophen concentrations in the third and fourth ventricle (lumped as TFV) as well as in the subarachnoid space (SAS), the latter being most representative of the lumbar CSF concentrations [135]. Denotations: In the model clearance (CL, volume/time), and ECF bulk or CSF flow (Q, volume/time) are indicated. Numbering indicates exchange between different compartments: 12 from plasma to peripheral compartment; 21 from peripheral to plasma compartment; 13 from plasma to brain ECF compartment; 31 from brain ECF to plasma compartment; 14 from plasma to CSF_LV compartment; 41 from CSF_LV to plasma compartment; 15 from plasma to CSF_TFV compartment; 51 from CSF_TFV to plasma compartment; 16 from plasma to CSF_CM compartment; and 61 from CSF_CM to plasma compartment.

Figure 5

Observed and predicted distribution of acetaminophen in human brain. a) The human physiologically-based pharmacokinetic model which equals the rat physiologically-based pharmacokinetic model, but includes human instead of rat physiological parameters. (For the denotations in the model see Figure 4b). b) Acetaminophen concentrations in human plasma and brain. Data points represent observed data in human for plasma (black diamonds) and lumbar CSF (orange circles) by Bannwarth et al.[137]. Lines represent predictions of human plasma concentrations (black line), human lumbar CSF concentrations (orange line, and human brain ECF concentrations (blue line) by the “humanized” preclinical physiologically-based PK model [135].

Figure 6

Brain distribution of quinidine in the rat [Westerhout J, Smeets J, Danhof M, De Lange ECM: The impact of P-gp functionality on non-steady state relationships between CSF and brain extracellular fluid. J Pharmacokin Pharmacodyn, submitted]. Average (geometric mean ±SEM) unbound quinidine concentration-time profiles following: a) 10 mg/kg, with co-administration of vehicle (-); b) 20 mg/kg, with co-administration of vehicle (-); c) 10 mg/kg with co-administration of 15 mg/kg tariquidar (+), and d) 20 mg/kg with co-administration of 15 mg/kg tariquidar (+). Black, blue, green and red symbols represent plasma, brain ECF, lateral ventricle CSF and cisterna magna CSF, respectively. Open symbols indicate data obtained without (-) and closed symbols represent data obtained with (+) the P-gp blocker tariquidar, respectively. The data show substantially lower concentrations in brain ECF (striatum) compared to lateral ventricle and cisterna magna CSF concentrations for both the 10 and 20 mg/kg dose (a, b). Upon co-administration of tariquidar, the brain ECF (striatum) concentrations were higher than those in the CSF compartments (c, d). These data show that the relationship between brain ECF and CSF concentrations is influenced by P-gp-mediated transport.

Figure 7

Brain distribution of remoxipride (REM) in the rat following intravenous (IV) and intranasal (IN) administration. Observed data points for plasma and brain ECF concentrations in the rat following intranasal and intravenous administration of remoxipride (open circles), and the “visual predictive check (VPC)” of the median concentration predictions of the model (black line), and the 90% prediction intervals (grey area). The VPC indicated that the model adequately described the observed data (from [147] with permission).

Figure 8

Observed and model prediction of remoxipride concentrations in human plasma (from [148,151,165] with permission). Human data on remoxipride and prolactin plasma concentrations were obtained following double intravenous administration of remoxipride at different time intervals. Data points on remoxipride concentrations in plasma (y-axis, concentration of remoxipride in μmol/L) as a function of time (x-axis, time in hours) are presented for each individual human subject (open symbols, DV). Using allometric scaling the preclinical PK model of remoxipride was tuned to the human PK model. The human PK model successfully predicted the remoxipride plasma kinetics in humans: the individual prediction of the median remoxipride concentrations is indicated (IPRE, ^________) as well as the population prediction (PRED, ---------).