ABSTRACTS 79

nonhuman primates as graft recipients. Clinical trials have also been conducted employing either fetal transplants or adrenal autografts for the treatment of parkinsonism. This presentation examines some of the issues which need to be considered in assessing the feasibility of using transplantation therapy for the clinical treatment of Alzheimer's disease. The results of recent studies from our group on nonhuman primates are also reviewed. As a starting point, there are several possible strategies for using neural transplants which could be considered based on our present understanding of Alzheimer's disease. Studies in rodents have utilized tissue grafts into the forebrain to augment deficient neurotransmitters such as acetylcholine. Another approach em- ployed in rodents has been to supplement the level of neurotrophic factors, especially Nerve Growth Factor (NGF), which is known to stimulate the basal forebrain cholinergic neurons that degenerate in Alzheimer's disease. We have preliminary studies examining both of these approaches in nonhuman primates. In a recent experiment, four young adult Cebus apella monkeys with unilat- eral nucleus basalis lesions received implants of fetal basal forebrain neurons into the cortex and amygdala. In the two monkeys receiving the youngest donor tissue (basal forebrain tissue from a 60-day embryonic Cebus donor) numerous grafted magnocellular cholinergic neurons were observed organized in clusters similar to those observed in situ. Long processes emanated from these cells, but did not cross the graft host interface. Functional studies were not done on these animals, but the results are a first step in demonstrating that fetal cholinergic neurons survived well following grafting into the primate forebrain. The second set of experiments (as part of a larger study on parkinso- nian monkeys) examined the feasibility of using autografted Schwann cells as a source of NGF in the primate brain since Schwann cells in transected peripheral nerve synthesize NGF. Three aged hemiparkinsonian Rhesus monkeys received peripheral nerve (sural nerve) autografts into the striatum. Three months later, numerous surviving Schwann cells were identified in the transplant site by routine staining procedures as well as staining for NGF receptor. These results suggest that peripheral nerve implants can serve as a species-specific source of NGF in the brain for at least limited periods of time. It has been proposed that neural transplantation techniques may some day prove useful for the treatment of Alzheimer's disease, but the basic scientific investi- gations needed to substantiate this proposal are in their early stages. Rodent studies from a number of laboratories, including ours, have been encouraging. The first experiments on nonhuman primates are now underway. Before clinical trials can even be considered, it is essential that systematic studies continue using appropriate animal model systems in order to assess the benefits as well as the potential risk of this mode of therapy. (Study supported by NIH grants NS25778, NS25655 and the American Health Assistance Foundation.)

POLYMERIC DELIVERY OF DRUGS TO REGIONAL BRAIN STRUCTURES: ANATOMIC AND BEHAVIORAL CONSE- QUENCES. M. Scan Grady, Matthew A. Howard and Marc R. Mayberg. Department of Neurological Surgery, University of Washington, Seattle, WA.

Pharmacologic therapy aimed at increasing CNS acetylcholine activity in patients with Alzheimer's disease has been disappoint- ing, perhaps because of nonspecific or subtherapeutic drug admin- istration. An alternative method of drug delivery exists, using polyanhydride polymer microspheres that are loaded with the desired drug and injected directly into the brain parenchyma. We tested polymeric drug delivery on a series of rats after performing

bilateral fimbria-fornix axotomy to cholinergicaily denervate the hippocampus in a simulation of the loss of cholinergic output seen in Alzheimer's disease. The first study examined twenty rats who underwent bilateral fimbria-fornix axotomy and two weeks later received bilateral hippocampal implants of bethanechol-impreg- nated polyanhydride polymer, blank polymer, or saline-only injections. For 40 days following the second procedure the animals underwent spatial memory testing daily using an eight-ann radial- maze. Animals receiving bethanechol polymer showed significant improvement by 10 days postimplant and improvement was maintained for the entire trial period. The two control groups failed to show any improvement. Histology confirmed complete- ness of axotomy and l~lacement of polymer. The second study examined diffusion of °H acetylcholine (ACh) from the injection site in the hippocampus. At four intervals over a period of 40 days, rats that received hippocampal injections of 3H ACh were sacri- ficed and brain sections submitted for autoradiography. A linear release of ACh was found, the label diminished to 50% at 40 days. The maximum spread of labelled ACh was 2 mm from the implant. There was very little gliotic reaction. This same experi- ment has been repeated using fluorescein-labelled dextran of molecular weight 10K, 20K, 40K or 70K injected into the cortex. The 10K dextran was retrogradely transported by neurons. The other dextrans showed a somewhat wider diffusion pattern. Poly- anhydride polymers may be a useful medium for localized delivery of molecules or drugs for short-term requirements.

POTENTIAL USE OF CNS DIRECTED DRUG DELIVERY SYSTEMS IN ALZHEIMER'S DISEASE. Robert E. Harbaugh. Section of Neurosurgery, Dartmouth-Hitchcock Medical Center, Hanover, NH.

Increasing knowledge of the neurochemical aspects of central nervous system function raises the possibility of treating Alz- heimer's disease (AD) and other neurological diseases by the appropriate manipulation of neurotransmitters, neuromodulators, neurohormones or neurotrophic factors. Clinical application of this knowledge may, however, be inhibited by long standing problems with drug delivery to the central nervous system (CNS) such as adverse systemic effects of neuroactive compounds, peripheral metabolism of drugs, inadequate blood-brain barrier penetration, erratic drug absorption, serum protein binding and poor patient compliance. The use of implantable drug infusion pumps, implant- able polymer drug delivery systems and neural tissue transplanta- tion to deliver drugs directly to the cerebrospinai fluid (CSF) or brain parenchyma may overcome many of these problems. Each of these delivery systems is evaluated with respect to its ability to reliably deliver compounds to a specific region or regions of the CNS, its applicability to a wide range of compounds and its degree of risk to the patient. These techniques are still in their infancy but are feasible and offer numerous potential advantages for drug delivery to the brain in the treatment of neurological diseases.

POTENTIAL USE OF NERVE GROWTH FACTOR IN THE TREATMENT OF ALZHEIMER'S DISEASE. F. Hefti. Andrus Gerontology Center and Department of Biological Sciences, University of Southern California, Los Angeles, CA 90089.

NGF is a well characterized protein which supports the matu- ration and the maintenance of viability and function of some populations of neurons. NGF is synthesized and released in the target region of responsive neurons and elicits its actions by binding to specific receptors on their surfaces. Alzheimer's disease is associated with pronounced loss of a group of NGF-responsive