Pharmacokinetics of antiepileptic drugs in conscious, mobile rats based on repeated, serial sampling of blood and CSF

Most studies on the pharmacokinetics of drugs have included single time point samples of blood or cerebrospinal fluid, often obtained under stressful conditions from large numbers of animals. There are few studies where serial, simultaneous sampling of blood and CSF has been carried out in fully conscious, mobile animals. In humans, although practically easier, such studies cannot be carried out for ethical reasons. This thesis has focused on the development of a model for serial simultaneous sampling of blood and CSF in conscious, mobile and stress-free rats, to be used for studying peripheral and central pharmacokinetics of phenytoin and sodium valproate. This work has comprised: (i) Establishing the model for serial, simultaneous blood and CSF sampling in conscious, fully mobile rats; (ii) Comparing the survival, causes of death, cannula patency rate and practical factors associated with the experimental procedures in two rat strains; (iii) Examining the metabolic effects of the surgical interventions and sampling procedures in the animals; (iv) Applying the model for pharmacokinetic studies of phenytoin and sodium valproate by serial blood and CSF sampling following once only (acute) and during five days (chronic) of drug administration at three and two daily doses respectively, to give serum concentrations relating to the therapeutic concentrations quoted in humans. The study has demonstrated that: (i) The model is associated with high survival and cannula patency rates, provided a rat strain with a low anaesthesia-mortality rate is used. (ii) The metabolic consequences of the experimental procedures are minimal and not likely to significantly affect pharmacokinetics of the drugs to be investigated. (iii) The pharmacokinetics of phenytoin confirmed dose and duration of dosage dependent phenytoin pharmacokinetics, but the results also showed that the transport into the CSF seems rate-limiting and non-linear, increasing disproportionately little with high doses and that except at steady state, the CSF levels do not reflect free or total serum phenytoin levels. At lower doses CSF concentrations of phenytoin were higher than the free fraction in serum, suggesting larger amounts of protein-bound drug in the CSF at low serum phenytoin concentrations than was noted before. (iv) The pharmacokinetics of sodium valproate confirmed dose dependency. The CSF levels included significantly more valproate than can be accounted for by the free fraction alone, particularly during non-steady state, and did not reflect total sodium valproate levels in serum except at steady-sate. The results suggested that significant amounts of CSF valproate may be derived from the serum protein-bound fraction but unlike with the phenytoin, there was no concentration-dependent barrier to the transport of valproate into the CSF