It is patently obvious that CNS activity (and trans-cellular permeability in general) is a complex function of physical/chemical properties of molecules such as size, lipophilicity, hydrogen-bonding potential, charge, and conformation. For any given molecule, one of these factors may dominate others. Drugs with the brain as the site of action should, in general, be able to cross the BBB. Drug delivery to the brain can be enhanced by increasing the lipophilicity of the molecule, by using prodrugs that dissociate after crossing the BBB, or by using passive or active drug targeting that utilizes transport systems at the BBB in the normal or disease states. In general, the trans-endothelial transport of compounds can depend on binding to constituents of the plasma, ionization state, time-dependent plasma concentration, and cerebral flow. It is possible to modify many of these properties with changes in chemical structure.
Previous attempts at understanding CNS activity have resulted in certain rules-of-thumb. For example, Andrews et al. have shown that an aromatic ring-tertiary nitrogen pharmacophore is important for CNS activity. Levin has successfully correlated octanol/water partition coefficient (LogP) and brain capillary permeability for compounds with molecular weight less than 400. However other, more recent attempts conclude that the octanol/water partition coefficient does not correlate well with blood-brain transport. Other criteria, like a limit of 8-10 hydrogen bonding groups per molecule, have also been proposed.