Pharmacokinetic-pharmacodynamic modeling of CNS drug effects: an overview

The use of kinetic-dynamic interactions in the evaluation of drugs

Deterred by the complexity of the mathematics, pharmacologists and clinical pharmacologists have only recently appreciated the usefulness of pharmacokinetics in drug development. Now unfortunately, although the vernacular of the science is known, often the meaning behind the words is lost. It is often assumed that drug levels are linearly related to drug action. Frequently they are not. This review shows, with reference to psychotropic drugs, how, in simple terms, it is possible to relate pharmacokinetics with pharmacodynamics, and how such relationships may provide a greater insight into drug activity and enhance drug development. Assuming that an equilibrium exists between the drug in plasma levels, and at the site of action, the same Michaelis-Menten equations used to relate effect to drug receptor binding can be used for drug level-dynamic interactions. A number of these relationships have been published and are discussed in terms of their derivation and their limitations. The graphical and computerised methods to create complete Emax curves are described and how the parameters of maximal effect, potency, variability and slope can be measured. When the drug is not in equilibrium with its site of action, hysteresis occurs and drug levels are out of phase with activity. Anticlockwise hysteresis, that is, activity increasing with time for a given drug level, can be caused by uptake into an active site, active metabolites, cascade activity, and sensitisation whilst clockwise hysteresis, in which activity decreases with time, can be caused by tolerance, active antagonistic metabolites, learning effects and feedback regulation. Attempts to relate simultaneously kinetics and dynamics by Link models can be difficult and not always necessary. It is assumed in therapeutic drug monitoring that individuals will show the same response for a given drug level. On the contrary, differences in individual subject sensitivity to drugs measured by kinetic-dynamic relationships may provide a greater understanding of the disease itself.

Actions and interactions of psychotropic drugs on human performance and mood: single doses of ORG 3770, amitriptyline, and diazepam

Actions and interactions of two antidepressants and diazepam on human skilled performance and mood were studied in a randomized double-blind cross-over trial with single oral doses of 50 mg amitriptyline (AMI), 15 mg Org 3770 (ORG) and placebo, given alone and in combination with 15 mg diazepam (DZ) to 12 young healthy subjects at one-week intervals. Objective tests (digit substitution, tapping, flicker fusion, Maddox wing, tracking, choice reactions, body sway, memory) and subjective assessments (visual analogue scale) were performed at baseline and 1.5, 3, 4.5 and 6 hours after drug administration. Side-effects were reported, blood pressure and heart rate measured and blood samples taken after each testing run. Placebo was nearly inert on performance and mood. DZ impaired some objective skills and showed sedative effects in the subjective tests. AMI produced sedation and impaired coordination as well as cognitive performance (digit substitution), most clearly at 3 to 4.5 hr. ORG resembled AMI in impairing objective and subjective performance, however, not necessarily in the same tests. Their combined effects with DZ were additive in objective tests but less additive in subjective tests. The drug combinations, but not any single drug, impaired learning acquisition. Plasma concentrations of the drugs given alone were about as expected, without important interactions. We conclude that the combinations of benzodiazepines with the antidepressants used impair skilled performance but may not cause major hazards.

Relationship between receptor occupancy at 37 degrees C and the anticonvulsant effect of flunitrazepam in rats

In this investigation an attempt was made to evaluate quantitatively the relationship between benzodiazepine receptor occupancy and the anticonvulsant effect of flunitrazepam in rats. A graded measure of anticonvulsant effect was obtained on the basis of an elevation of pentylenetetrazol (PTZ) threshold concentrations. The concentration-anticonvulsant effect relationship could be described by the Emax model with an EC50 in cerebrospinal fluid of 2.9 +/- 0.8 micrograms/liter and an Emax of 227 +/- 22 mg/liter PTZ (mean +/- SE). In vitro receptor occupancy was determined in a crude brain homogenate at 0 and 37 degrees C, which yielded KD values of 2.2 +/- 0.2 and 26 +/- 2 micrograms/liter, respectively. The results obtained in both experiments were combined by focusing on free flunitrazepam concentrations. This strategy resulted in a nonlinear relationship between receptor occupancy and anticonvulsant effect of flunitrazepam, with 90% of the maximum response achieved at a degree of receptor occupancy of approximately 50% at 37 degrees C.

Strategy to assess the role of (inter)active metabolites in pharmacodynamic studies in-vivo: a model study with heptabarbital

The purpose of this investigation was to develop a universal experimental strategy by which the role of (inter)active metabolites in in-vivo pharmacodynamic studies can be examined. Heptabarbital was chosen as a model drug and several pharmacokinetic variables which may affect in-vivo concentration-pharmacological response relationships were examined. Adult female rats received an i.v. infusion of the drug at one of three different rates (0.225-1.50 mg min-1) until the animals lost their righting reflex (after 11 +/- 1 to 88 +/- 8 min of infusion). The serum concentration of the drug at onset of loss of righting reflex (LRR) increased slightly with increasing infusion rate. The drug concentrations in brain tissue and cerebrospinal fluid (CSF), (mean +/- s.d.: 67 +/- 5 mg kg-1 and 24 +/- 4 mg L-1, respectively, for the lowest infusion rate) were not affected by the infusion rate. The possible contribution of (inter)active metabolites to the pharmacological response of heptabarbital was determined by administration of different i.v. bolus doses (14.1-22.5 mg) resulting in widely differing sleeping-times (7 +/- 3 to 119 +/- 20 min). The concentrations of heptabarbital in serum, brain tissue and CSF at offset of LRR (mean +/- s.d.: 77 +/- 8 mg L-1, 76 +/- 7 mg kg-1 and 29 +/- 5 mg L-1, respectively, for the highest dose) were not affected by the administered dose.(ABSTRACT TRUNCATED AT 250 WORDS)

Application of serial sampling of cerebrospinal fluid in pharmacodynamic studies with a drug active in the CNS: heptabarbital concentrations at onset and offset of loss of righting reflex in rats

As cerebrospinal fluid (CSF) possesses unique characteristics in order to explore concentration-pharmacological response relationships of drugs active in the CNS, the practicability of serial sampling of CSF was tested in a study with heptabarbital. Concentrations in CSF and plasma were measured simultaneously in individual rats during and after an intravenous infusion for 30 min. At the end of the infusion, the distribution equilibrium was attained with a CSF/plasma concentration ratio of 0.38, roughly equal to the fraction unbound to protein. When concentrations in blood and CSF were determined at the onset and offset of loss of righting reflex concentrations in blood were significantly greater at onset (146 +/- 19 mg/l) than at offset (108 +/- 16 mg/l, n = 6), whereas concentrations in CSF were identical (39 +/- 5 and 38 +/- 5 mg/l, respectively). This confirmed the earlier observation that the CSF is pharmacokinetically indistinguishable from the site of action. When the duration of the loss of righting reflex was varied, concentrations of heptabarbital in CSF at onset and offset were similar, independent of the duration of the loss of righting reflex (1-5 hr). These findings demonstrate the absence of the development of acute tolerance and confirmed that no (inter)active metabolites interfered with the pharmacological response. In a total number of 26 rats the concentrations in CSF at onset and offset of loss of the righting reflex were compared. The interindividual variation was 13-15% and the intra-individual variation was only 4-6%. The results demonstrate the usefulness of serial sampling of CSF in pharmacodynamic studies with centrally acting drugs.