Intracranial controlled release polymers have been used for drug delivery to the brain, bypassing the blood brain barrier (BBB). By understanding the rates and patterns of transport in the local tissues, it is possible to design delivery systems that provide the optimal spatial and temporal pattern of chemotherapy within the intracranial space. This paper reviews the kinetics of drug release from polymeric controlled release implants, and describes the fate of drug molecules following release into the brain interstitium. Potential improvements in drug delivery based on the understanding of the mechanisms of drug release, transport and elimination are discussed.

We have studied the release of nerve growth factor (NGF), a protein under consideration for treatment of Alzheimer's Disease, from polymer matrices and microspheres to characterize the stability of NGF, the dynamics of NGF release, and the distribution of NGF within the brain interstitium. Poly(ethylene-co-vinyl acetate) (EVAc) disks and poly(L-lactic acid) (PLA) microspheres were formed by codispersing NGF with one of a variety of molecules. The mass of mouse NGF (mNGF) detected following release from EVAc disks into buffered saline varied five-fold over the range of codispersants studied, with carboxymethyldextran providing optimal release, while the mass of recombinant human NGF (rhNGF) released varied four-fold from both EVAc disks and PLA microspheres, with albumin and carboxymethyldextran providing optimal release. Variation of the codispersant species significantly affected NGF release into buffered saline; it also had a noticeable, but small, effect of the amount of NGF found in the brain tissue following implantation of a polymer device. To improve NGF retention in tissue, NGF was conjugated to 70 000 molecular weight dextran and incorporated into a polymeric device. The distribution of NGF was enhanced by conjugation; comparison of NGF concentrations in the brain to a mathematical model of diffusion and elimination suggested that the elimination rate of NGF-dextran conjugate in the tissue was over seven times slower than the elimination rate of NGF. These results indicate that variation of the properties of the controlled release system may be useful in regulating the time course of NGF delivery to tissue, and that modification of the NGF itself can improve penetration and retention in the brain.