Central Nervous System (CNS) Penetration

Central nervous system penetration is a quantity, not a yes-or-no property. The modern expression of it is the unbound partition coefficient: the ratio of free drug concentration in brain interstitial fluid to free concentration in plasma. A value near one means the compound distributes passively to equilibrium, well below one means net efflux is winning, and above one means an active uptake transporter is concentrating it. Only free drug engages a receptor, so this ratio, rather than total brain content, predicts a central effect.

Measuring it properly is laborious. Microdialysis samples interstitial fluid directly in animals, brain homogenate binding experiments correct total tissue concentration for nonspecific lipid binding, and cerebrospinal fluid sampling is the usual human proxy. The strongest human evidence is positron emission tomography showing displaceable binding at the intended receptor. For most therapeutic peptides the honest answer is that unbound penetration is very low, and the successful central peptide programmes have worked around that rather than through it.

The number separates two explanations for the same observation. If a compound produces a central effect while its unbound brain exposure is negligible, the mechanism is peripheral by elimination: afferent nerve signalling, circumventricular access, or a downstream mediator that does cross.

The classic misreading is a total brain-to-plasma ratio from a homogenate. A lipophilic base can look highly penetrant by that measure while being almost entirely bound to membranes and unavailable, and unperfused tissue counts drug still sitting in cerebral blood. Intracerebroventricular administration in rodents is also cited as evidence of penetration when it shows the opposite: the route exists precisely because the compound does not get in on its own.