Pharmacokinetics of intraventricular administration.

The widespread use of neurosurgical devices for intraventricular drug delivery has led to an escalation in the number of CSF pharmacokinetic studies. When experimental data for ventricular CSF concentration vs. time are normalized for dose and plotted for drugs with a wide range of physical properties, there is a remarkably narrow range. Simulations from a distributed model for the central nervous system suggest that the narrow range of observed CSF half-times following bolus administration is a consequence of physiologic limits upon rate of drug removal from CSF. Whereas bulk flow of CSF establishes minimum rate of drug washout, diffusion through brain tissue and subsequent removal by capillaries establishes a maximum rate of drug egress which is only seven times the minimum rate. Predictions of CSF concentration vs. time can be made based upon the capillary exchange rate for a particular drug, which could be estimated or calculated from CSF concentration measurements following systemic administration. Simulations are also presented of steady-state CSF concentrations during continuous ventricular infusions.