Skip secondary menu

Parkinson's Disease: A Research Planning Workshop

Table of Content

 Executive Summary

Virtually all Americans will be touched by Parkinson's disease in some way, whether they develop the disease themselves or know someone else who is afflicted. This disease impairs control of movement, progressing from symptoms such as tremor and muscular rigidity to total disability and death. At present, the cause of Parkinson's is unknown and there is no cure. The purpose of the Parkinson's Disease Research Planning Workshop was to bring together Parkinson's disease researchers and experts from other fields to foster new ideas and research efforts that might lead to rapid advances in understanding and treating the disease. The workshop was organized around the themes of etiology, pathogenesis, and therapy, summarized below.


  • The proportion of inherited Parkinson's cases is not yet clear. Some people do appear to inherit Parkinson's disease, and several large multi-case families have been carefully studied. Regardless of the percentage of inherited cases, finding genes responsible for familial Parkinson's should be helpful for understanding all forms of the disease. Techniques now available should allow researchers to find the genes responsible for familial Parkinson's disease in a relatively short time.
  • Some toxins and viruses can cause Parkinson's-like disorders. Population studies suggest that the environment may play a role in the development of some cases of Parkinson's disease. In spite of considerable research effort, however, no environmental factor has yet been implicated in the majority of Parkinson's cases. A large ongoing study is now examining possible interactions between environmental agents and genetic susceptibility.
  • Age is the most important risk factor for Parkinson's disease. While some age-related changes are similar to the degeneration of Parkinson's disease, the disease is not simply a result of normal aging. Age-related processes may add to brain degeneration from Parkinson's disease to produce the signs and symptoms of the disease.


  • Biomarkers are measurable character-istics that indicate disease risk, presence of early and unsymptomatic disease, or disease progression. While several biological abnormalities have been found in Parkinson's patients, there is a critical need for reliable, affordable, and non-invasive biomarkers to help understand the disease and evaluate potential new therapies. Researchers also need to develop a clear definition of Parkinson's disease, which is one of several disorders with similar signs and symptoms. This is especially important for evaluating therapies, which may affect various Parkinson's-like disorders differently.


  • The cause of Parkinson's disease is unknown. The death of nerve cells that use dopamine as a neurotransmitter in the substantia nigra is responsible for most of the motor control problems, but other cells also are affected. Damage to these other cells may contribute to the resistance of some Parkinson's signs and symptoms to dopamine-based therapies.
  • There is increasing recognition that similar biological mechanisms contribute to many neurodegenerative diseases. Several of these interacting mechanisms have been specifically implicated in the pathogenesis of Parkinson's disease. Among them are oxidative stress, mitochondrial defects, cytoskeletal abnormalities, excitotoxicity, calcium-mediated damage, and programmed cell death.
  • The realization that cells help regulate their own fate and may invoke "suicide" programs under some circumstances is an important insight into degenerative diseases. Better understanding of these regulatory programs may offer targets for therapeutic intervention in Parkinson's disease.
  • Trophic factors, known to be important in development, are now recognized to participate in maintaining a healthy nervous system in adults. Some trophic factors have shown promise as therapies in animal models of Parkinson's disease. However, improved understanding of the basic biology of these factors in the brain is needed. Moreover, practical problems must be overcome before trophic factors will be useful for human patients.
  • While several interacting biological processes have been implicated in Parkinson's disease, the trigger for the disease is still unknown. Finding genes for familial Parkinson's disease may provide important clues about the underlying cause.


  • Standard medical therapy alleviates symptoms by using the dopamine precursor levodopa to replenish the depleted neurotransmitter dopamine. However, this therapy does not stop the underlying degeneration of brain cells.
  • Loss of dopamine neurons in the substantia nigra disrupts motor control circuits in the brain of Parkinson's patients. In particular, those pathways that inhibit movement may dominate the pathways that activate movement. Surgical techniques to remove parts of the brain or implant chronic stimulating electrodes aim to restore the balance between movement-inhibiting and movement-activating circuits. Electrical stimulation has the advantage of being reversible but may require recurrent adjustments and present other practical problems.
  • Surgical destruction of brain tissue, as in pallidotomy, was tried with inconsistent results before the advent of levodopa therapy. Now surgeons are re-examining this procedure using the behavior of individual brain cells, recorded with microelectrodes, as signposts to guide them to precise locations in the brain. Results have been promising and systematic trials are now underway.
  • The promising results of modern surgical approaches reflect the benefits of improved understanding of the neurophysiology of motor control. The results demonstrate that therapies need not concentrate on replacing dopamine to be effective. Questions raised by these therapies highlight the need for better understanding of how the brain's circuits control movement.
  • Attempts to replace dopamine cells by transplantation of fetal tissue show that transplanted cells can survive for long periods and benefit some patients. In these patients, transplantation has reduced, but not eliminated, the need for levodopa. Transplantation methods in the future may use cultured cell lines and stem cells rather than fetal tissue and will require better understanding of trophic factors and other cell signaling molecules.
  • Gene therapy using transplantation of genetically engineered cells or direct insertion of genes into patients' brain cells may become clinically useful in the long term. Basic research is needed to develop better vectors for gene insertion, to improve long-term gene expression, to define the best target cells for gene therapy, and to find genes other than those related to dopamine that might be useful in treating the disease.

The workshop highlighted several avenues through which research efforts may help people afflicted with Parkinson's. However, Parkinson's is a complex disease and there is no definitive cure on the immediate horizon. Progress is likely to occur in steps and to require efforts on several basic and clinical fronts. Improved understanding of the underlying biology of the disease will lead to better ways of relieving the symptoms of Parkinson's patients and ultimately halting the underlying degeneration of brain cells.


 Parkinson's Disease: A Research Planning Workshop

America has entered a "golden age" of neuroscience research. Advances in basic understanding of the nervous system are providing opportunities to develop therapies for previously untreatable diseases of the brain. In recognition of these advances, Congress designated the 1990s the Decade of the Brain and mandated a research emphasis on neurological diseases of concern to many Americans. In 1995, the midpoint of this decade, Congress encouraged the National Institutes of Health to sponsor a research planning workshop on Parkinson's disease. This debilitating disease affects more than 500,000 Americans and causes progressive symptoms including tremor, muscle rigidity, and immobility that ultimately lead to total disability and death. The total annual direct and indirect cost of Parkinson's disease was estimated to be $6 billion in 1992. Because the disease most commonly affects people in later life, the number of people with Parkinson's disease and the associated costs will grow as the average age of the American population increases.

The cause of Parkinson's disease is unknown and there is no cure. The major site of degeneration in the brain is the substantia nigra, where neurons producing the neurotransmitter dopamine progressively die. At present, most people with Parkinson's disease receive drugs designed to replace or mimic dopamine in the brain.

While these drugs are initially effective in most patients, they do not stop the underlying neurodegeneration. In addition, they often have unpleasant side effects which become progressively more troublesome, causing patients to relinquish their independence and productivity.

The Parkinson's Research Planning Workshop, co-sponsored by the National Institute of Neurological Disorders and Stroke, the National Institute on Aging, the National Institute of Environmental Health Sciences, and the National Institute of Mental Health, took place August 28-30, 1995, at the Madison Hotel in Washington, D.C. The workshop's purpose was to bring together key Parkinson's disease researchers and experts from other fields to foster new ideas and research directions that might lead to rapid advances in the understanding and treatment of the disease.

The workshop planners asked participants to suggest fundamentally new ideas rather than strategies for achieving incremental advances. Recent progress in a variety of research fields is now yielding insights that might lead to more effective therapies or even ways of preventing Parkinson's disease. These advances include sophisticated new methods of manipulating molecules and cells; an enhanced understanding of the brain's motor circuits; knowledge of growth factors and their functions in the brain; rapid advances in the molecular genetics of human disease; and a growing understanding of the common ways in which neurons respond to stress and injury, regardless of the regions of the brain or the organism in which they are found. The workshop, organized around the topics of etiology, pathogenesis, and therapy, was designed to find ways of using this increased knowledge to more rapidly advance research in Parkinson's disease.

The discussion centered upon several major themes. Chief among these was the recognition that both genetic and environmental factors are important in understanding Parkinson's disease. Many participants also emphasized a need to identify biological traits, or biomarkers, that would allow researchers to identify people at risk for developing Parkinson's disease, allow earlier diagnosis of the disease, and mark its progression. Participants encouraged collaboration between basic and clinical scientists. In recognition of the common themes emerging from research on different organisms and diseases, they also called for collaboration with scientists outside the Parkinson's field.



Virtually all neurodegenerative diseases arise from the interaction of genetic predisposition, environmental factors, and time, especially aging. In the 1980's, a dramatic finding that the unusual toxicity of the chemical MPTP mimicked many aspects of Parkinson's disease focused attention on environmental toxins that might trigger the disease itself. Clinicians have noted familial tendencies in Parkinson's disease for about a century, although only in the last several years has increased evidence brought recognition that genetics indeed plays a role in the disease. This represents a major change in thinking from 5 or 10 years ago, when genetic factors were not considered important. Aging is clearly the most important risk factor for Parkinson's disease, as for many other disorders. Understanding how genetics, environment, and aging combine to produce Parkinson's disease may provide vital clues about pathogenesis and treatment.


Substantial clinical evidence now demonstrates the importance of genetic factors in Parkinson's disease. The growing list of multiple-case families with clinically and pathologically documented Parkinson's disease, together with the frequency of familial history among patients in clinics, demonstrate an influence of heredity. The accumulated clinical and family history information and the power of modern molecular technology argue for research aimed at finding the responsible genes.

Familial (or hereditary) Parkinson's disease may be much more common than has been thought in the past. Some clinicians believe that about 10-15 percent of all Parkinson's disease is familial. A recent study examined the family histories of 216 patients diagnosed with Parkinson's disease at one clinic in a given period. The geneticist found full pedigree information for first and second degree relatives (including all four grandparents) in 19% of these cases. Among patients in this group, more than half so far have other close family members with confirmed or probable Parkinson's disease. There is still uncertainty about what proportion of Parkinson's disease is familial. However, the proportion may be as high as in Alzheimer's disease and amyotrophic lateral sclerosis, two neurodegenerative diseases in which genes have been identified in the familial forms of the disease. Even if a genetic form of the disease is uncommon, discovery of the gene responsible would help in understanding the sporadic form. The recent highly publicized genetic findings for breast cancer illustrate this, as have investigations into amyotrophic lateral sclerosis and Alzheimer's disease.

At least eleven families with multiple cases of Parkinson's disease have been studied extensively. Some are modest in size, but others have more than 20 confirmed cases. For example, four generations of clinicians at the Mayo Clinic have followed one midwestern family. The founder of this family died in 1913 with a death certificate indicating paralysis agitans (Parkinsonism); autopsy material from other family members examined as early as the 1920's is still available. With 23 affected members now identified, the full expression of the disease in this pedigree - including considerable variability of signs and symptoms - is coming to light. In Italian-American and Greek-American pedigrees, disease occurrence appears similar in family members who stayed in Europe and those who migrated to America. In addition to the families with classical Parkinson's disease, several related hereditary syndromes have been identified, with and without the pathological Lewy body inclusions that are characteristic of Parkinson's disease. This suggests that more than one gene might be involved. These family pedigrees represent the accumulation of many years of clinical and genetic research and could be a valuable resource for identifying the genes underlying Parkinson's disease. Whether most Parkinson's patients, or only a small subset, inherit the disease, efforts to identify the genes responsible are now feasible.

Rapid progress in molecular genetics has made it easier to find genes contributing to disease. There are three strategies for finding a disease gene. In the traditional approach, researchers begin with knowledge of the primary functional deficit responsible for a disease, isolate the defective protein, and proceed to identify and map the gene for that protein. The genes for Tay-Sachs disease and phenylketonuria were found in this manner. Because the primary biochemical deficit responsible for Parkinson's disease remains elusive, this is not presently an attractive approach. The positional strategy for finding a gene begins with family pedigrees and linkage analysis and proceeds through examination of genetic markers to identify the chromosomal position and ultimately the responsible gene. With gene in hand, the tools of modern molecular biology may uncover the role of the gene product in the disease. This approach has succeeded in identifying genes for cystic fibrosis, ataxia-telangiectasia, and for the familial Alzheimer's chromosome 14 gene, and may be directly applicable to Parkinson's disease. The third approach, which might be called inspirational, begins with an educated guess based on known pathophysiology and known genes; that is, scientists nominate candidate genes and either exclude or confirm them using information about inheritance. The familial Alzheimer's chromosome 21 gene was found in this way, as were genes for those forms of retinitis pigmentosum resulting from mutations in rhodopsin. This strategy becomes more powerful as researchers identify more genes. However, since nearly half of the roughly 100,000 human genes participate in nervous system function, a chance success in neurological disease is unlikely. Several candidates for Parkinson's disease, mostly genes for enzymes of dopamine metabolism, have been excluded using this approach. Calculations suggest that the largest three Parkinson's pedigrees are each sufficient to identify a disease-related gene using the positional approach and available technology.

While optimism about finding Parkinson's genes is certainly warranted, several factors can hamper genetic analysis. Among these is incomplete penetrance ü the failure of a genetic disease to show overt symptoms in persons who have the underlying genetic defect. Parkinson's disease is often not evident until late in life, and people often die due to other causes before the disease symptoms appear. Another potential obstacle is the presence of more than one gene that can independently cause a disease. If families in which disease results from different genes are lumped together for analysis, the relationship of particular genes to the disease can be obscured. Such genetic heterogeneity is clearly the case for familial Alzheimer's disease, and is not unlikely for Parkinson's disease given the apparent complexity of the pathogenesis and the existence of several clinically different forms of familial Parkinsonism. Finally, polygenic determination can make genetic identification particularly difficult. In this mode of inheritance, many genes interact to produce a disease, but no single gene is responsible. The asymmetry of Parkinson's disease in different sides of ancestral trees argues against polygenic determination, however. Geneticists have methods to cope with all of these problems when sufficient data is available. Some pedigrees for familial Parkinson's disease are growing large enough that a gene search within a single family may be possible, sidestepping potential complications such as the presence of different disease-related genes in different families.



Scientists have long suspected a role of environmental factors such as toxins or viral infections in development of Parkinson's disease. This theory gained impetus in the early 1980s, when researchers reported Parkinson's-like symptoms and neuron loss in many people exposed to the designer drug MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) through intravenous drug abuse. This finding showed that Parkinson's-like symptoms can result from exposure to toxins. Many scientists now believe most Parkinson's disease could result from an interaction between environmental factors and genetics.

Much of the evidence for environmental factors in Parkinson's etiology comes from population studies. One such study found that the risk of acquiring Parkinson's disease in the United States rose slightly from 1935 to 1985. However, the statistical significance of this rise is uncertain and it may reflect improved diagnosis rather than increased incidence of disease. The relative risk of Parkinson's disease also is generally higher in the U.S. and other industrialized countries than in less industrialized ones. These findings suggest that an unidentified industry-related exposure, such as a toxic chemical, might lead to Parkinson's in certain people, possibly those with increased genetic susceptibility. However, despite intensive investigations, researchers have not found any environmental factors that can account for the majority of Parkinson's cases. Some occupational exposure studies have found that farmers and other agricultural workers have an increased risk of developing Parkinson's disease, suggesting that agricultural chemicals may play a role in the disease pathogenesis. This link is supported by studies linking Parkinson's to dieldrin and dithiocarbamate pesticides. Interestingly, the toxin MPTP is structurally similar to certain pesticides. Immediate Parkinson's-like symptoms also developed in one individual who ingested a mixture of petroleum products. While the results are intriguing, the evidence is preliminary and requires more study.

Other evidence implicating toxins in Parkinson's disease comes from a unique combination of Parkinsonian symptoms, dementia, and motor neuron disease found in many people from the island of Guam. Some researchers believe this disease might result from a toxin found in cycad seeds, which were eaten in greater amounts during times of food shortage. This toxin mimics the excitatory neurotransmitter glutamate.

Several findings suggest that viral infection also may lead to Parkinson's or similar diseases. Many people who contracted viral encephalopathy during an influenza epidemic in the early 1900s later developed severe and progressive Parkinson's-like symptoms. Similar symptoms developed in a group of Taiwanese women shortly after herpesvirus infections. Although in the latter instance these symptoms appeared linked to temporary inflammation of the brain's substantia nigra and later disappeared almost completely, they showed that herpesvirus can affect the region of the brain associated with Parkinson's disease.

Many questions remain about the role of environmental factors in Parkinson's disease, and how they may interact with genetic influences. Some of the questions may be answered by twin studies, especially a large study of twins who are veterans of World War II. This study is examining disease rates in fraternal and identical twins and their exposures to some environmental factors. Since this study includes over 20,000 people, researchers can avoid some problems of earlier twin studies of Parkinsonism, which yielded inconclusive results because they lacked statistical power. In addition, studying Parkinson's cases that overlap with other neurodegenerative diseases, such as Alzheimer's disease, may provide insights about how known genes and environmental exposures may affect Parkinson's disease.


Aging is the most important risk factor for Parkinson's disease. The frequency of Parkinson's increases markedly with age in people older than 50. Recent progress in understanding normal aging has yielded insights about how it may interact with the pathogenic processes of Parkinson's disease and has provided clues to understanding the variable expression of the disease.

Death of dopaminergic neurons in the substantia nigra is a central aspect of the pathogenesis of Parkinson's disease, and some of the same cell groups show attrition in normal aging. However, less than 20 percent of nigral dopamine cells die in normal aging, whereas in Parkinson's disease more than 60 percent of these cells may die before the disease becomes clinically evident. More extensive cell loss occurs as the disease progresses. The pattern is similar for other markers of aging that are enhanced in disease. Changes associated with normal aging may push Parkinson's patients closer to the threshold for symptom expression, but Parkinson's disease is not simply normal aging.

The dramatic effects of diet restriction on aging provide clues about how aging might exacerbate Parkinson's disease. Restricting food intake of rodents by about 20 percent increases their lifespans 30 to 40 percent, and all their organ systems exhibit a slowed aging schedule. Diet restriction also blunts some features of basal ganglia aging, such as the amphetamine effect on behavioral rotation in rats; this is of interest because basal ganglia malfunction is critical in Parkinson's disease. (These very restrictive rodent diets are radical and experimental and there is no evidence that a similar strategy would help humans with Parkinson's disease.) The full spectrum of mechanisms by which diet restriction slows aging is not yet understood, but a reduction in oxidative damage, which seems to be a major contributor to the pathogenesis of Parkinson's disease, may play a role. Disruption of the cytoskeleton and defects in mitochondrial function also are common to both aging and Parkinson's disease. In addition, aging processes involving glial cells and the dopamine-degrading enzyme monoamine oxidase might contribute to greater production of endogenous neurotoxins in Parkinson's disease.

There is considerable variation in the expression of familial Parkinson's disease and in the Parkinson's-like signs and symptoms of people exposed to the neurotoxin MPTP. The study of development - of which aging is a part - has revealed that chance is another source of individual variability, apart from genetics or environment. Unlike the strictly determined course in some invertebrates, chance is important in determining cell fate during mammalian development. Data have shown a two-fold range of cell numbers in the brainstem's locus ceruleus and in the spinal roots of even highly inbred rodents. Cell counts in other regions of the rodent brain and in other mammals suggest that such individual variations are likely to be a widespread phenomenon. Chance developmental differences in the numbers of reserve neurons in affected brain areas might account for some of the variation in how and when Parkinson's disease appears.


Genetics, environment, and aging all may contribute to the development of Parkinson's disease. While continued research into the contributions of environment and aging is clearly necessary, recent advances in molecular genetics could allow genes that influence Parkinson's disease to be found in a relatively short time. Identifying those genes is particularly important for progress in understanding the pathogenesis of the disease and finding a cure.



A central theme of the workshop was the need to develop reliable, affordable, and non-invasive biomarkers for use in predicting and diagnosing Parkinson's disease and measuring its progression. Biomarkers are biological characteristics used to indicate or measure disease risk, presence of disease, and disease progression. Workshop participants agreed that reliable biomarkers are critically needed and that identifying them would yield dramatic benefits. Clinicians can rate disease progression using rating scales and other strategies, but no practical biomarkers have yet been identified.

Many potential biomarkers have been proposed, including specific differences in cellular activity, gene defects, manifestations of underlying abnormalities, and cellular characteristics. While many of these proposed biomarkers show great promise, most are untested and based upon characteristics only tentatively linked to Parkinson's disease. Many also are invasive or present other practical problems for widespread clinical use.

Risk-associated biomarkers help identify individuals at high risk for a disease. People at risk can then be monitored for preclinical signs of disease and be targeted for prevention studies. One proposed risk-associated biomarker for Parkinson's disease is a specific form of a gene called CYP2D6, which breaks down toxins in the substantia nigra. One study found that people with the L form of CYP2D6 on both chromosomes have more than a five-fold increase in the risk of developing Parkinson's disease. Because Parkinson's is a complex disease, a combination of genetic risk factors may be needed to predict the likelihood of developing disease. Other proposed preclinical biomarkers include abnormalities of mitochondrial complex 1 respiratory enzymes; abnormalities of a dopamine-degrading enzyme called monoamine oxidase B (MAO B); and abnormalities of the cytochrome P450 enzyme, needed for detoxifying certain alkaloids.

Preclinical diagnostic markers are essential for trials intending to halt progression of Parkinson's disease because extensive degeneration always occurs before symptoms appear. Any attempt to halt progression of the disease should therefore begin as soon as possible. Valid and reliable preclinical diagnostic markers would help reduce the costs of clinical trials by preventing enrollment of patients who are unlikely to benefit. Such markers also would allow design of clinical trials designed to prevent Parkinson's symptoms from appearing. Proposed preclinical diagnostic markers include measures of dopamine deficiency in the brain's substantia nigra and changes in dopamine uptake in the nigrostriatum. Experimental PET and SPECT brain scans have shown promise for detecting these changes in dopamine systems, but more study is needed. Other potential markers include neuropsychological tests; ability to identify odors; altered glutathione levels; abnormal reflexes; and abnormal mitochondria in blood platelets. Much more research is needed, however, to evaluate whether any of these proposed preclinical markers is really useful.

Developing improved disease progression markers would be a boon to clinical researchers because they need to determine whether therapies affect just symptoms or the underlying progression of the disease. Suggested progression markers for Parkinson's disease include brain scans measuring neurodegeneration, measurements of dopamine metabolite levels in cerebrospinal fluid, or elimination of the need for levodopa. Improved markers would allow more focused clinical trials and reduce the associated costs.

Despite intensive studies, no biomarkers have been proven suitable for widespread clinical use in Parkinson's disease. Progress in basic understanding of pathogenesis may be needed to identify practical biomarkers. Once identified, potential biomarkers require substantial testing for reliability and validity. However, advances in this area could lead to more rapid improvements in treating Parkinson's disease and reduced expenses in clinical trials.

Researchers also need to develop a clear definition of Parkinson's disease, which is one of several disorders with similar clinical signs and symptoms. This is particularly important for evaluating new therapies, which may affect various Parkinson's-like disorders differently. As a result of the workshop recommendations, the National Institute of Neurological Disorders and Stroke has undertaken an effort to develop refined guidelines for Parkinson's diagnosis.



The cause of Parkinson's disease is unknown. Current treatments relieve symptoms but do not halt the progression of the disease. Today's standard medical therapies arose in the 1960s from the recognition that degeneration of dopaminergic neurons projecting from the substantia nigra to the striatum causes most of the motor problems. There is still no clear answer about what causes these neurons - and neurons in other parts of the brain - to degenerate. However, progress in cell biology is illuminating common processes that damage cells and mechanisms by which cells respond to damage and regulate their own life and death. Researchers are linking these general themes to the specific pathogenesis of Parkinson's disease. They must learn how these processes influence the death of specific groups of neurons, produce the chronic and progressive character of the disease, and disturb the brain circuitry to cause signs and symptoms.

 Neuronal Degeneration, Lewy Bodies, and Cytoskeletal Proteins

The degeneration of dopaminergic neurons of the substantia nigra is central to current understanding of Parkinson's disease. Surviving neurons in the substantia nigra often contain Lewy bodies, abnormal structures that take up particular biological stains. Lewy bodies are a pathological hallmark of Parkinson's disease. Their presence in other areas of the brain, particularly brainstem areas like the locus ceruleus that send processes throughout the brain, indicates that Parkinson's disease is not just a disease of nigral dopamine cells. In diffuse Lewy body disease, a type of dementia defined in 1980, Lewy bodies appear throughout the brainstem and cerebral cortex. How best to distinguish Parkinson's disease from other related syndromes like diffuse Lewy body disease is not yet clear, but several parts of the nervous system in addition to the substantia nigra are affected in the latter disorders. The involvement of regions in addition to the substantia nigra probably underlies the resistance to treatment of some motor symptoms such as speaking difficulties, neck rigidity, postural problems, and gait disorders; of non-motor symptoms such as cognitive and mood disturbances; and of autonomic nervous system dysfunction. Degeneration in these brain areas also may account for the inability of drugs to suppress many symptoms during late disease progression.

Immunohistochemical staining methods have revealed that neurofilament proteins are major constituents of Lewy bodies. Neurofilaments are important components of the neuronal cytoskeleton. Neurofibrillary tangles, a prominent pathological finding in Alzheimer's disease, contain neurofilament proteins, but unlike Lewy bodies they also contain tau protein. The small protein ubiquitin is a sensitive marker for Lewy bodies. Because tagging with ubiquitin is a widespread process by which cells condemn proteins for rapid degradation, often in response to damage, Lewy bodies may reflect a response to cellular damage. Whether the presence of cytoskeletal proteins in Lewy bodies reflects a primary defect in the cytoskeleton or a secondary consequence of damage, as by oxidative stress, is not yet clear. A primary defect in neurofilament structure or regulation could lead to retraction of neuronal processes and cell death. In fact, transgenic mice producing excessive amounts of a neurofilament protein have cellular structures similar to Lewy bodies that lead to the death of some neurons.


In excitotoxicity, excessive release of normal excitatory neurotransmitters damages neurons. Glutamate is the most prevalent excitatory transmitter in the central nervous system, and glutamate excitotoxicity clearly contributes to the acute damage caused by stroke and trauma. Recent evidence also implicates excitotoxic processes in neurodegenerative disorders, such as amyotrophic lateral sclerosis, and suggests how excitotoxicity might produce the selective neuronal degeneration and progressive time course characteristic of Parkinson's disease.

The epidemiological circumstances surrounding the unique combination of motor neuron, Parkinson's-like, and dementia symptoms in Guam present one line of evidence that excitotoxicity can produce a chronic disorder. Cycad seeds, consumed in greater amounts during times of food scarcity, may cause this disorder even though symptoms often do not arise until years after the seeds are eaten. The toxic factor in these seeds powerfully affects glutamate receptors.

Excitotoxicity might cause specific patterns of neuronal damage via differential sensitivity to glutamate, susceptibility to the downstream consequences of excessive excitation, or distribution of excitotoxic agents other than glutamate. Several different combinations of glutamate receptor subunits can form functional receptors. Scientists are just beginning to understand how subunit composition differs among brain areas and controls receptor characteristics. The complement of receptors in some brain areas might predispose them to excitotoxic damage. While scenarios for specificity based on glutamate receptor subtype distributions are speculative, there is solid evidence that neurons in the substantia nigra are particularly susceptible to oxidative damage, one of the main downstream consequences of excessive excitation. Finally, neurons in the substantia nigra harbor high concentrations of dopamine, and oxidation products of dopamine can be toxic. A slow leakage of dopamine from nigral cells leading to increased production of these toxins might tip the balance from normal excitation towards toxicity.

 Oxidative Stress, Energy Metabolism, and Mitochondrial Dysfunction

One mechanism by which excitotoxicity damages cells is oxidative stress. Oxidative damage by free radicals is a major contributor to normal aging and to many diseases, especially degenerative diseases.

Free radicals are highly reactive molecules formed during normal oxidative metabolism and by other normal and pathological cellular processes. Each of these molecules has an orbital with an unpaired electron and aggressively seeks to steal an electron from another molecule. When a free radical captures an electron (i.e. oxidizes another molecule), it may create another radical and so initiate a damaging chain reaction. Many cellular components are susceptible to oxidative damage, including DNA, lipids, and cytoskeletal components. The extent of oxidative stress reflects a dynamic balance between the forces that promote oxidation and the critical defense systems the body has evolved to protect cells from damage. Even a slight shift in that balance can eventually yield dire consequences.

The brain is especially vulnerable to oxidative stress. Neurons are rich in polyunsaturated lipids, which are targets for oxidation. The brain also consumes a disproportionately large share of the body's oxygen intake and so creates many free radicals. Iron levels are high in the brain, and iron promotes reactions that create free radicals. The brain has less of some antioxidant defenses than other body tissues, and so may be less able to defend itself. Finally, of course, the central nervous system has less capacity for self-repair than other body organs.

The substantia nigra is among the brain regions most vulnerable to oxidative stress, and oxidative damage in the nigra is evident in Parkinson's disease. Nigral neurons contain melanin, which selectively binds iron and other metals that promote the formation of free radicals. The increased iron, along with normal levels of the iron-binding protein ferritin, in the substantia nigra of Parkinson's patients indicates that iron is in a potentially harmful reactive form. Nigral cells also, of course, harbor high levels of dopamine. As dopamine undergoes metabolism it generates hydrogen peroxide. The reducing agent glutathione normally clears the peroxide, preventing harm, but hydrogen peroxide reactions promoted by iron may produce highly reactive and dangerous hydroxyl radicals. The diminished levels of glutathione and other defensive antioxidants reported in Parkinson's disease may exacerbate this problem. Finally, Parkinson's disease patients have elevated markers for oxidative damage of lipids and DNA and abnormal mitochondrial function.

Mitochondria are the powerhouses of the cell. They extract useful energy by carefully controlled, stepwise oxidation (essentially slow burning) of carbohydrates. For this reason, enzymes of the inner mitochondrial membrane produce 95 percent of the free radicals in the body. Mitochondria are therefore both a source of damaging radicals and front line targets for damage, providing the potential for a vicious cycle. Primary mitochondrial defects often occur in severe infantile or juvenile diseases, but there are now clear examples of these defects causing diseases that arise later in life. The primary site of MPTP's toxic action appears to be within a functional unit of mitochondria called complex I. While the presence of complex I defects in Parkinson's disease itself had been disputed in the past, perhaps because of the technical difficulties of necessary experiments, there is now more convincing evidence for defects of complex I in the substantia nigra of Parkinson's patients. There also are reports of other mitochondrial abnormalities. The normal age changes in mitochondrial function might interact with these disease specific changes to promote Parkinson's disease.

The complex I mitochondrial defect may be a primary defect in Parkinson's or may be secondary to damage from internal or external agents. Mitochondrial defects are the primary cause of several diseases. Because mitochondria carry their own DNA, inherited only from the mother, mitochondrial diseases may follow a maternal pattern of inheritance. The patterns of occurrence of some mitochondrial diseases illustrate how environmental and hereditary factors can interact to cause disease. For example, the aminoglycoside class of antibiotics has well-known potential to cause deafness; in separate Shanghai and Israeli-Arab populations a maternally inherited mitochondrial abnormality causes an increase in antibiotic binding and increased vulnerability to these antibiotics' side effects. Thus the genetic predisposition and the toxic insult interact to cause deafness in many of these people, although excessive levels of antibiotics alone can produce deafness even without an inherited predisposition. Although maternal inheritance of mitochondrial defects can occur in some diseases, the cell nucleus carries most of the genes for mitochondrial proteins. Therefore, mitochondrial defects usually follow normal patterns of inheritance.


 Calcium and Nitric Oxide

In addition to their role in energy metabolism, mitochondria play a crucial role in calcium regulation by actively buffering large changes of calcium within cells. Cells must tightly manage internal calcium concentrations because calcium levels control crucial enzyme systems, cytoskeletal dynamics, neurotransmitter release, and other critical aspects of cellular biochemistry. Most excitotoxicity results from uncontrolled increases in cellular calcium concentration.

One surprising agent of calcium-induced damage is the simple gas nitric oxide (not to be confused with the anesthetic nitrous oxide). This small molecule, which can diffuse through cell membranes, is a signal used by neurons, epithelial cells, and immune cells. In neurons, for example, nitric oxide may participate in long-term changes in synaptic strength responsible for learning, and in vascular epithelial cells it acts as a signal to control blood vessel diameter. The enzyme nitric oxide synthase (NOS) synthesizes nitric oxide. Calcium governs the activity of this enzyme through the regulator molecule calmodulin. So, in one sequence of excitotoxic events, high levels of glutamate activate the NMDA subtype of glutamate receptor, allowing calcium to flow into cells. Calcium then binds to calmodulin, which in turn activates NOS. Activated NOS produces excess nitric oxide, which combines with oxygen radicals to form the extremely reactive free radical peroxynitrite. This potent oxidizing agent damages cellular constituents and often leads to cell death.

There is compelling data that nitric oxide is an important agent of neuronal damage in animal stroke models. Mice genetically engineered to lack the neuronal form of NOS, that is NOS "knockout" mice, show much less damage following cerebral ischemia. Inhibitors of NOS also can minimize damage of cultured neurons by excitotoxins. Nitric oxide may be a mediator of damage in Parkinson's disease as well.

An enzyme inhibitor selective for the neuronal form of NOS can reduce the Parkinson's-like effects of the toxin MPTP in rodents. Nitric oxide also interferes with the mitochondrial enzymes and affects iron metabolism, both of which may contribute to oxidative damage pathways of Parkinson's disease.

 Trophic Factors and Cytokines

Nerve growth factor (NGF), discovered some fifty years ago, controls the survival and differentiation of developing peripheral sensory and autonomic neurons. In the last few years, a newly recognized role for NGF and other trophic factors in the central nervous system has been proposed. Scientists are demonstrating that trophic factors are important in development of the nervous system, in maintenance of the nervous system in adults, and probably in the response to injury and aging.

Several new trophic factors have been discovered recently, and many more probably exist. The most interesting for Parkinson's disease is GDNF. This peptide was isolated and identified from a glial cell line (hence the name glial-cell-line-derived-neurotrophic factor), leading to the cloning of the gene. GDNF was first shown to support dopamine neurons in cell culture. Subsequently, 6-hydroxydopamine and MPTP models of Parkinson's disease in rodents demonstrated the potency of this factor in vivo. For example, 6-hydroxydopamine injection into the median forebrain bundle of the rat produces unilateral damage to the pathway connecting the substantia nigra to the basal ganglia. GDNF injected into the substantia nigra one month after such a toxic lesion was shown to profoundly reduce the behavioral effects of damage. Other biochemical and histological markers supported the conclusion that GDNF saved nigral cells and promoted useful growth of neuronal processes. The extent of recovery progressively increased for several weeks after the GDNF injection.

One surprise from these studies was evidence that many of the rescued cells survived the initial toxic insult but changed phenotype, that is, they stopped making the enzymes responsible for dopamine synthesis. MPTP toxic models of Parkinson's disease have produced similar results.

The efficacy of GDNF underscores the importance of trophic factors, but the details of the GDNF story reveal crucial gaps in our knowledge about the basic biology of these molecules. The intensively studied biology of NGF serves as a model for all growth factor research. Appropriate targets for innervation by NGF sensitive neurons release very small amounts of NGF. Specific receptors on axon terminals of NGF sensitive cells bind NGF, internalize it, and carry it by axonal transport back to the cell body. There NGF acts through defined biochemical signaling pathways to control gene expression in the cell nucleus. Defects in many of these processes, such as in the production of trophic factors, in receptors, in cytoskeletal mediated axonal transport, or in biochemical signaling pathways, might plausibly underlie a disease like Parkinson's. For GDNF, however, most of these aspects, especially receptors and signaling pathways, are poorly understood. What is more, circumstantial evidence suggests that GDNF may not be the biologic factor most important for dopamine neurons or for other neurons affected by Parkinson's disease. Peptide signaling molecules, including GDNF, are members of extended families of structurally similar molecules, and GDNF may be acting on dopamine neurons because of its resemblance to a more appropriate sibling factor. Better understanding of the basic biology of GDNF is necessary to predict whether Parkinson's patients will respond to GDNF like the toxin-injected animal models do, and how and where in the brain a trophic factor might best be administered. Intensive investigations to find new trophic factors, to characterize their receptors, to delineate biochemical signaling pathways, to understand the regulation of receptor and factor expression, and to elucidate the physiologic roles of trophic factor signaling in adult and developing brain are certainly warranted.

The GDNF family of peptide signaling molecules includes several important cytokines. Cytokines, originally described as growth, differentiation, and activity controlling factors for immune and other blood cells, are found in healthy brains and can have effects on neurons and glia. For example, they help regulate expression of transmitters and neuropeptides and participate in processes that control synaptic strength.

Following injury and during some diseases, cytokine levels rise dramatically in the nervous system. Cytokine inflammation mediators may also participate in the pathology of chronic neurodegenerative diseases. Epidemiological evidence suggests that agents which block inflammation (the non-steroidal anti-inflammatory agents widely used as pain relievers) may slow the onset of Alzheimer's disease. In Parkinson's disease there is evidence for activated microglia in the substantia nigra, possibly reflecting local cytokine activity. Several cytokines may be present in the striatum but not in the cerebral cortex during Parkinson's disease, and some subsets of immune cells, such as gamma delta T cells, may be higher in the blood of Parkinson's patients. However, cytokine regulatory systems are extremely complex, and the role of cytokines in the onset and progression of Parkinson's disease is not yet clear.


 Programmed Cell Death

Researchers are only beginning to understand the interrelationships among the various factors implicated in Parkinson's disease. Every neuron is thought to integrate information about damage to DNA, the cytoskeleton, proteins, and lipids, to check toxin, calcium, and virus levels, and to factor in levels of trophic substances, cytokines, and hormones. If appropriate, the cell initiates a program for self-repair or self-destruction. The realization that cells actively help regulate their own fate is an important theme of modern cell biology, with many implications for medicine. Failure to activate programmed cell death may result, for example, in autoimmune disease, the proliferation of tumors, or establishment of latent viral infections, while excessive programmed death may contribute to developmental disorders or neurodegenerative diseases.

The term apoptosis refers to the characteristic structural changes in cells undergoing programmed death and often connotes more generally the entire process. In apoptosis, the cell membrane undergoes certain changes and membrane enclosed fragments of the cell may break away; the nucleus condenses and fragments; and polyribosomes break up. In most cells, enzymes are activated to cut DNA into pieces that appear as a "DNA ladder" on biochemists' analytical gels and serve as attachment points for histochemical markers. The DNA degradation may be a defense against viruses that have attempted to establish residence. Finally, dying cells induce surrounding cells, microglia, or macrophages to engulf and scavenge the debris. The essential consequence of apoptosis is that damaged cells eliminate themselves without releasing intracellular enzymes and dangerous small molecules like dopamine and glutamate that might harm their neighbors. The controlled process of apoptosis contrasts with the more corrosive effects of necrosis. In necrosis, cell membranes break down, cells release their contents, and inflammation ensues.

Apoptosis is believed to occur in the substantia nigra of patients with Parkinson's disease. Many dopamine cells in the substantia nigra of Parkinson's patients have been reported to meet the structural criteria for apoptosis in electron micrographs, while few cells, if any, appear apoptotic in normal control brains. Dopamine neurons also display cytokine receptors that can trigger apoptosis and express the bcl-2 gene product, an important regulator of the cell death program.

Whether most dopamine neurons die from apoptosis or necrosis, or whether apoptosis in Parkinson's disease reflects defective cell death control systems or an appropriate response to cellular damage, the fact that cells can turn on programs to speed or retard their own death is an important concept in neurodegeneration. Several proteins controlling cell death programs have been identified, but much about the control systems remains unknown. Control proteins discovered in experimentally accessible organisms like nematode worms show remarkable similarity to the control proteins of mammals. This evolutionary conservation of function testifies to the fundamental importance of these pathways and facilitates progress in understanding mammalian systems. Viruses also trick cells into giving them safe harbor by using molecules that mimic the cell's control factors to suppress apoptosis. These "trojan horse" control factors have proven useful in research and hold promise for therapeutic intervention as well. Control systems for cell death offer sites for intervention in neurodegenerative disease even if they do not participate directly in the pathology of the disease.


Several pathogenic processes contribute to Parkinson's disease. Researchers can devise two-hit and multi-hit disease models, analogous to those employed to explain carcinogenesis, to help understand how the effects of genetics, the environment, and aging result in Parkinson's disease. However, the starting point of the disease process is unknown. The use of knockout mice lacking specific biochemical pathways and transgenic mice with excessive expression of known enzymes could be a powerful strategy to help resolve the issue. Biomarkers specific for particular pathogenic processes might also help. A likely path for determining the initial cause of Parkinson's disease is by identifying the genes responsible for familial Parkinson's disease.

Even before the trigger and course of the disease process are fully understood, progress in understanding Parkinson's disease can suggest several approaches to therapeutic intervention. Trials based on some of these common disease processes, including antioxidants, trophic factors, and glutamate antagonists, are already underway for other diseases. Other therapeutics based on these insights into disease processes may require a substantial investment in basic biology before they can be applied.



Standard therapy for Parkinson's disease consists primarily of administering the drug levodopa, a precursor of dopamine. Levodopa often is combined with other agents to enhance its effect. Other drugs are available, but do not work as well as levodopa in reducing Parkinson's symptoms. All current Parkinson's drugs can have side effects, including dyskinesia - uncontrolled, often continuous movements. None of the currently available drugs stops the underlying degeneration associated with Parkinson's, and the effects of drug therapy often wear off over time. Researchers are now experimenting with a number of advanced surgical and non-surgical approaches to treating Parkinson's. They hope these new therapies will help patients who do not benefit from current drugs and will perhaps even slow the course of the disease.

 Palliative Therapies

Current levodopa therapy for Parkinson's disease stems from the premise that dopamine depletion is responsible for the signs and symptoms of Parkinson's. Levodopa is initially very good at relieving symptoms in most patients, however it becomes less effective with time, usually with an increase in adverse effects. Researchers are studying whether effectiveness of levodopa and other drug therapy for Parkinson's can be improved by altering the method and timing of drug administration.

While treating Parkinson's disease with levodopa reduces many symptoms, it does not restore the brain to normality. Many changes besides dopamine neuron loss occur in the brain as a result of Parkinson's disease, and these may complicate the treatment. A variety of drugs, which affect other neurotransmitter systems such as acetylcholine or GABA systems, can be used with levodopa therapy to try to compensate for some of these changes. However, these drugs have other complications and are less effective than levodopa when administered alone.

Some of the major side effects of levodopa therapy are associated with fluctuations in dopamine levels in the brain, and thus in the degree of symptomatic response. In addition, these unnatural fluctuations in brain dopamine may contribute to the appearance of levodopa-induced dyskinesias. Changes in blood pressure and cognition also tend to fluctuate with levels of dopamine. Researchers are experimenting with a number of methods to stabilize levels of the drug. These include slow-release polymers, drug pumps, patches, and other controlled release strategies; COMT (catechol-o-methyltransferase) inhibitors, which extend the response to levodopa by slowing degradation; and drugs similar to levodopa but with different pharmacokinetics (rates of uptake into and elimination from the body). Ironically, some studies suggest that con-tinuous administration of levodopa con-tributes to development of drug tolerance, so continuous infusion may not be ideal.

A critical question regarding levodopa therapy is what role it plays in its own eventual failure to help those with Parkinson's disease. Some animal studies imply that levodopa may in fact be toxic to dopamine neurons and that its use ultimately contributes to progression of the disease. This is a critical and controversial issue because large numbers of people are currently using levodopa. Researchers need to learn whether the quantity of the drug or the method of administering it contributes to levodopa failure. Another factor is that continued loss of dopamine neurons or other brain cells contributes to the altered drug response. Finally, more must be learned about how the normal dopamine system functions and how precursors and relatives of dopamine function in the brain once neurons and their connections degenerate.

Participants emphasized the importance of developing better receptor-activating drugs for dopamine and other neurotransmitter systems. Finding ways to improve current drug therapy will lead to immediate benefits for Parkinson's patients and further improve scientific understanding of the disease.


Among the experimental therapies for Parkinson's disease, researchers are testing strategies that preserve or restore function and integrity to vulnerable neurons and slow or halt illness. These therapies can take many forms and can begin either before or after the disease develops. Developing neuroprotective therapies hinges upon developing rational interventions based on pathogenesis and etiology and on finding effective techniques for delivering the therapy to the brain. Learning which neurons to target and identifying effective biological markers of disease risk and progression are also important.

Many intervention strategies have been suggested for Parkinson's disease based on the known pathogenesis of the disease. Only two - tocopherol (an antioxidant) and deprenyl (a monoamine oxidase B inhibitor) have been systematically tested, however. Other potential therapies include trophic factors, especially BDNF, GDNF, EGF, and bFGF. While these trophic factors appear promising in laboratory studies, difficulties with drug administration have slowed their use in clinical trials. Other potentially neuroprotective agents include remacemide (an epilepsy drug with suspected neuroprotective effects) and glutamate receptor blockers (to protect against excitotoxicity). They also include COMT inhibitors that prevent breakdown of dopamine and drugs that act on dopamine transporters and receptors.

Tocopherol and deprenyl have been tested in slowing progression of Parkinson's disease through a multi-center, randomized, controlled clinical trial known as the DATATOP (Deprenyl and Tocopherol Antioxidative Therapy of Parkinsonism) trial. Tocopherol was not beneficial during the initial 2-year period, but patients receiving deprenyl were 50 percent less likely to need levodopa therapy to maintain independence than control patients. A follow-up study recently showed, however, that this benefit was not sustained and subjects required levodopa therapy during the 24-42 months following the initial study. The Parkinson's Study Group, which conducted the DATATOP trial, continues to study deprenyl and has begun studying other therapies for Parkinson's, including COMT inhibitors and clozapine (an antipsychotic drug) for Parkinson's patients with psychosis. Other potential neuroprotective therapies must be tested in a similar manner to show whether they are in fact beneficial.

Workshop participants again emphasized the importance of identifying and validating biomarkers. They also advocated carefully assessing which neuroprotective measures are most likely to be effective.


 Genetic Engineering

Among the new therapies being developed for Parkinson's disease is gene therapy using genetically engineered cells or viral vector injections to add functioning genes to the brain. Gene therapy techniques also are being studied for several other neurodegenerative diseases, and they may be useful in repairing damage from stroke and other insults. Depending on the gene introduced, gene therapy for neurological diseases may restore missing neurotransmitters; halt degeneration of cells using growth factors; or repair damage due to a sudden event, such as spinal cord injury.

In Parkinson's disease, the main goal of gene therapy so far has been to increase levels of the neurotransmitter dopamine in the brain. Unlike standard drug therapy, gene therapy would provide a continuous, local supply of dopamine to specific brain regions. This might reduce side effects and problems associated with fluctuations of drug level in the body. However, each potential therapy must be carefully checked in animal models before being tested in humans.

The primary strategy for increasing dopamine through genetic engineering is to add the gene for tyrosine hydroxylase, an enzyme in the biochemical pathway that converts the amino acid tyrosine to dopamine. One method for doing this is to use a viral vector to insert the gene into skin cells or other cells taken from the patient, then insert the modified cells into the deficient brain region. This method has so far been tested only in animal models such as MPTP-injected monkeys. Using cells from the individual being treated eliminates rejection of the graft and the need to suppress the immune system.

Another method for gene therapy attempts to insert genes directly into brain cells. Viruses that have been modified to safely carry the TH gene are injected into the brain in the hope that the viruses will insert the gene into nearby brain cells in the process of infecting them. Eventually, scientists may find ways to insert foreign genes into cells through non-viral means, such as liposomes (artificial membrane enclosures used to deliver drugs or other substances into cells) or endocytosis (the process cells normally use to engulf foreign substances such as nutrients).

Successful gene therapy requires manipulation of many factors. Researchers must identify and insert the appropriate gene for the task. For cell transplant therapy, they need to insert this gene into cells that can be easily obtained, that are harmless to the host, and that can survive and provide stable, long-term gene expression in the host. In any gene therapy, scientists must find a safe and effective viral vector to insert the target gene into cells. They have experimented with several types of viruses, including retroviruses, herpesviruses, adenoviruses, and adeno-associated viruses. Each type of vector has certain advantages and disadvantages. Adeno-associated viruses, for example, are defective forms of human parvoviruses that are harmless in the human body. However, they require an inactivated "helper" virus, such as an adenovirus, to work in the body, and these "helper" viruses can cause cell death near the injection site. It also is uncertain whether adeno-associated viruses can yield long-term expression of the inserted gene.

Experiments using genetically altered skin and muscle cells in monkey models of Parkinson's disease have shown that the grafted cells can survive and produce tyrosine hydroxylase for several months, resulting in as much as 100 percent of the dopamine supply found in normal animals. They also have shown that the cells integrate well into the brain and do not migrate or cause death of surrounding cells. Researchers need to establish how many genetically engineered cells to insert for optimal dopamine production. They may eventually be able to improve gene expression by inserting better promoters - segments of gene-regulating DNA - with the target gene.

Continued research is critical for realizing the potential of genetic engineering. Ultimately, scientists may be able to insert genes into neural lineage stem cells instead of skin and muscle cells. Workshop participants emphasized the need to improve vectors for inserting genes, to increase the duration of gene expression, and to define good cell targets for future gene therapy.


Parkinson's researchers are testing several surgical therapies in clinical trials. Because these procedures are invasive, they are usually reserved for severely afflicted, late-stage Parkinson's patients in whom other treatments have failed.

Surgery was one of the earliest approaches used to treat Parkinson's disease. Pallidotomy (surgical destruction of a region of the globus pallidus), was very common in the 1950s and 1960s, but was abandoned almost completely after the introduction of levodopa therapy in the 1960s. When dopamine is lost in the brain's motor control systems, pathways that inhibit movement may overwhelm the pathways that activate movement. Pallidotomy and electrical stimulation aim to restore the balance of these movement-controlling systems. While early pallidotomies had a high complication rate and were not consistently effective, the procedure has now been resurrected using modern technology such as brain scans and microelectrode recording to precisely locate the region to be destroyed. With these painstaking measures, many recent pallidotomies have dramatically relieved the major symptoms of Parkinson's disease.

Pallidotomy is based on understanding of how motor control circuits are damaged in Parkinson's disease. The globus pallidus is part of the brain's basal ganglia region, which plays an important role in the pathway controlling movement. Normal motor control requires a critical balance between basal ganglia pathways that inhibit movement and those that promote it. The neurodegeneration in Parkinson's disease disrupts this balance, resulting in excessive inhibition of the cortical neurons that initiate movement. Changes in neuron activity in different parts of the basal ganglia have been well-studied in monkey MPTP models of Parkinsonism. The knowledge gained from these studies enables surgeons to use the behavior of individual neurons to precisely locate the region of the brain that needs to be destroyed.

Animal studies have shown that lesions of the subthalamic nucleus in MPTP-exposed monkeys can dramatically reduce Parkinson's-like symptoms on the side opposite each lesion. Inactivating the subthalamic nucleus by direct application of the drug muscimol, which mimics the inhibitory neurotransmitter GABA, has similar effects, suggesting that this region is important to production of Parkinson's symptoms. However, creating lesions in the subthalamic nucleus is dangerous because it has one of the densest concentrations of blood vessels in the brain, increasing the risk of hemorrhagic stroke. The globus pallidus, the next step in the pathway, is a safer and more clinically useful target.

Preliminary studies of microelectrode- guided pallidotomy show that all signs and symptoms of Parkinson's disease respond to pallidotomy to some extent. The wide variety of symptoms relieved by pallidotomy is difficult to reconcile with current models of basal ganglia function and suggests more complex functions within this brain region than described by theoretical models. Many patients who receive pallidotomy show bilateral improvement in symptoms even if only one side of the brain is lesioned. Patients receiving pallidotomy so far have shown no permanent dyskinesia or other motor impairments. A related procedure called thalamotomy (destruction of a part of the thalamus that participates in motor control) may be more effective in treating tremor, but pallidotomy is generally safer than thalamotomy, especially in older patients. Patients' age and location of the lesion affect the outcome of pallidotomy, and effects on motor symptoms are much more dramatic than effects on cognitive and psychological symptoms. Some patients do continue to deteriorate after the procedure, especially in the unlesioned side of the brain.

While apparently successful in the short-term, pallidotomy has several drawbacks, including partial loss of vision in about 11 percent of initial patients, potentially impaired cognition, and increased risk of hemorrhage and stroke. Careful monitoring and surgical expertise are necessary to reduce these risks. Pallidotomy also fails to stop the underlying disease progression, and patients usually need medication long-term.

Experiments with pallidotomy show that effective therapies for Parkinson's do not necessarily need to target the dopamine neurotransmitter system directly and suggest that current models of basal ganglia function are inadequate. A better understanding of the normal motor control pathways is badly needed. The findings also suggest that hypokinetic (slowed movement) and hyperkinetic (excessive movement) disorders result from opposite changes in activity of the basal ganglia and thalamocortical regions. These findings suggest many areas for future research. Scientists must now evaluate the long-term effectiveness of pallidotomy in treating symptoms and define how the effects of unilateral and bilateral lesions differ. A related surgical procedure called ansotomy, which consists of interrupting axon connections to the globus pallidus, also has been shown effective in animals and may soon be ready for testing in human trials.


 Electrical Brain Stimulation

High frequency brain stimulation is a reversible alternative to pallidotomy and thalamotomy. In this method, electrodes stimulate one of three brain structures: the ventral intermediate nucleus of the thalamus, the subthalamic nucleus, or the globus pallidus interna. Unlike surgical destruction of brain regions, this therapy is reversible since the electrode signal generators can be turned off.

In brain stimulation, surgeons use microelectrode recording in combination with stimulation to determine the best sites to implant the pulse-generating electrodes.

These implanted signal generators are coupled with radiofrequency-coupled pulse receivers. Studies of brain stimulation in several hundred patients have shown that complications are generally rare but can include hemorrhage, partial paralysis, and infection. Electrical stimulation also can have side effects such as speech slurring, abnormal touch sensations, and muscle contractions, depending on the brain region stimulated. Results of subthalamic nucleus stimulation are generally the best at reducing rigidity and akinesia (lack of movement), and subthalamic nucleus and thalamic stimulation are both very effective in reducing tremor. Unfortunately, the subthalamic nucleus is a particularly risky target because of its high concentration of blood vessels.

Electrical stimulation's main advantages compared to pallidotomy are its reversibility and adjustability. Complications usually occur only when the stimulation is turned on, and the stimulation can be tailored in individual patients to achieve the best response possible. Reversibility also allows easy, double-blind studies of the therapy's effects. Patients given electrical stimulation may still be able to benefit from newer therapies as they are developed. Stimulation also can be used in multiple brain regions, such as the subthalamic nucleus and the globus pallidus, or in combination with pallidotomy in the other brain hemisphere.

Electrical brain stimulation has several unique drawbacks. It requires a more elaborate initial surgical procedure, followed by additional procedures to change batteries. Adjustments and repair of broken electrodes also are sometimes required. Brain stimulation tends to be especially costly and may not work as well as pallidotomy in reducing symptoms. Finally, patients require "holidays" from the stimulation, and it is usually turned off at night to save batteries. This leads to symptom fluctuations much like those associated with levodopa therapy.

Brain stimulation may be improved with future research identifying other brain targets, and it also may prove useful in other diseases, such as accident-related tremor. Scientists also can study stimulation with multiple electrodes or in combination with drug infusions.


Another potential strategy for treating Parkinson's disease is transplantation of cells as a way of restoring dopamine transmission in the brain. Transplantation often incorporates genetically engineered cells to provide neurotrophic factors or other accessory molecules that help the cell grafts survive and grow. In animal models, cell grafts can improve growth and survival of damaged cells or build cell "bridges" that help neuron regeneration. In Parkinsonian patients, researchers have experimented with implants of fetal dopamine cells, adrenal cells, and immortalized cells.

The clinical results of transplantation experiments have so far been moderate, largely due to poor grafted cell survival rates (only 5 to 20 percent). However, the results also show that a long-term improvement of symptoms is possible, and some patients receiving the grafts have been able to reduce or even eliminate medication.

The best-tested transplantation strategy is grafting of fetal dopamine cells. In one study of more than 200 patients who have received dopamine grafts since 1987, the effects were moderate. After transplantation, most patients still required levodopa, but they developed a more prolonged response to the drug with better control of rigidity and hypokinesia. These symptoms improved rapidly, peaking anywhere from 3 months to 24 months following transplantation. Some grafted dopamine neurons remained alive in patients for more than six years. The grafted neurons connected only to nearby targets, so 90 to 95 percent of the putamen (a region of the basal ganglia) remained inactive. However, the study also showed that long-term immunosuppression to prevent rejection of the grafts is unnecessary and can be withdrawn after 6 months.

With current procedures, transplantation typically uses cells from three to four fetuses in each treated brain hemisphere. This raises ethical questions that may be resolved when it becomes possible to use cultured cell lines or bioengineered cells rather than fetal tissue. Researchers also may be able to improve survival of grafted cells, reducing the need for donor tissue. Animal studies have found that survival of cell grafts in rats improves five-fold if cells are co-grafted with skin cells genetically engineered to produce the trophic factor bFGF. Graft survival also improves three-fold with the addition of molecules called lazaroids.

In animal studies, scientists have found that transplantation works best if the host is young. Graft survival, growth, and connection to targets may be reduced with increasing age. Trophic factors that accumulate after cell damage and easy access to targets for axon growth can enhance graft survival. A good blood supply to the transplanted cells is also important. Other preliminary studies show that fetal and adult neuronal progenitor cells proliferate indefinitely in the laboratory when given neurotrophic factors such as bFGF. These propagated cells often survive if they are later grafted into the brain. Interestingly, adult neural progenitor cells grafted into the brain form glial cells instead of neurons if they are placed in an area without nerve connections. If placed in an undamaged area, however, they differentiate into normal neurons for that area.

Suggestions for future transplantation research include studying the optimal location of fetal grafts in relation to symptoms and individual patterns of neuronal loss. Researchers also must learn whether transplanting cells to more than one brain region might be beneficial. Finally, they need to substantially improve current transplantation techniques to find ways of improving graft survival and improving the reliability and reproducibility of the results. Transplantation methods in the future may use cultured cell lines and stem or progenitor cells rather than fetal tissue and may apply improved understanding of trophic factors and other cell signaling molecules.


Researchers are developing a variety of new surgical and non-surgical therapies for Parkinson's disease. These new therapies show great promise for improving the quality and length of life in Parkinson's patients. While some experimental therapies are already helping patients in clinical trials, scientists must find ways to refine these therapies and answer the questions surrounding their use. Other therapies, such as genetic engineering, require much more preclinical research before they can become available for trials. Improved understanding of Parkinson's disease and development of valid biomarkers for disease progression will aid study of these therapies and perhaps lead to new ones.



The Parkinson's Research Planning Workshop brought together scientists specializing in Parkinson's and other fields to generate new ideas and research strategies. Major themes included the contributions of genetics and the environment to development of the disease, the need for practical biomarkers and refined diagnostic guidelines for Parkinson's, the contribution of several cellular processes to the disease, and prospects for better therapies. While the discussion highlighted several key areas for investigation, Parkinson's is a complex disease and there is no definitive cure on the immediate horizon. Improved understanding of the underlying biology of the disease will lead to better ways of relieving the symptoms of Parkinson's patients and ultimately halting the underlying degeneration of brain cells.

Appendix: Presenters, Patient Representatives, and Organizers


Yves Agid, M.D., Ph.D.
Laboratoire de Medecine Experimentale

Krzysztos S. Bankiewicz, M.D.
Section Preclinical Studies
Somatix Therapy Corporation

M. Flint Beal, M.D.
Neurology Research

Alim L. Benabid, M.D., Ph.D.
Laboratoire de Neurobiophysique

Donald B. Calne, D.M., F.R.C.P.C.
Neurodegenerative Disorders Centre
Vancouver Hospital/Health Sciences Centre

Thomas N. Chase, M.D.
Experimental Therapeutics Branch
National Institute of Neurological Disorders and Stroke

Marie-Francoise Chesselet, M.D., Ph.D.
Department of Pharmacology
University of Pennsylvania

C.W. Cotman, Ph.D.
Institute for Brain Aging and Dementia
University of California, Irvine

Rex W. Cowdry, M.D.
National Institute of Mental Health
National Institutes of Health

Terri Damstra, Ph.D.
National Institute of Environmental Health Sciences
National Institutes of Health

Theodore M. Dawson, M.D., Ph.D.
Department of Neurology
The Johns Hopkins University School of Medicine

Mahlon R. Delong, M.D.
Department of Neurology
Emory University School of Medicine

D.W. Dickson, M.D.
Department of Pathology
Albert Einstein College of Medicine

Roger C. Duvoisin, M.D.
Department of Neurology
University of Medicine and Dentistry of New Jersey

Stanley Fahn, M.D.
H. Houston Merritt
College of Physicians and Surgeons
Neurological Institute

Howard J. Federoff, M.D., Ph.D.
Department of Neurology
University of Rochester Medical Center

Caleb E. Finch, Ph.D.
Neurobiology of Aging
University of Southern California
Ethel Percy Andrus Gerontology Center

Thomas B. Freeman, M.D.
Division of Neurosurgery
Department of Pharmacology and
Experimental Therapeutics
University of South Florida
College of Medicine

Fred H. Gage, Ph.D.
Laboratory of Genetics
The Salk Institute

Richard H. Haas, M.D.
Division of Pediatric Neurology
Department of Neurosciences
University of California at San Diego

Zach W. Hall, Ph.D.
Office of the Director
National Institute of Neurological Disorders and Stroke
National Institutes of Health

Franz Hefti, Ph.D.
Genentech, Inc.

Richard J. Hodes, M.D.
Office of the Director
National Institute on Aging
National Institutes of Health

Barry J. Hoffer, M.D., Ph.D.
Department of Pharmacology, C-236
University of Colorado Medical Center

Zaven S. Khachaturian, Ph.D.
Khachaturian, Radebaugh and Associates, Inc.
International Consultants on Alzheimers

Anthony E. Lang, M.D., F.R.C.P.
Division of Neurology
Movement Disorders Centre
The Toronto Hospital

J. William Langston, M.D.
The Parkinson's Institute

Virginia M.Y. Lee, M.D.
Department of Pathology and Laboratory Medicine
University of Pennsylvania School of Medicine

Olle F. Lindvall, M.D.
Department of Neurology
Lund University Hospital

Andres M. Lozano, M.D., Ph.D.
Department of Neurosurgery
The Toronto Hospital

Guy M. McKhann, M.D.
Zanvyl Krieger Mind/Brain Institute
The Johns Hopkins University

Robert L. Nussbaum, M.D.
Laboratory of Genetic Disease
National Center for Human Genome Research
National Institutes of Health

John G. Nutt, M.D.
Department of Neurology
University of Oregon

C. Warren Olanow, M.D.
Department of Neurology
Mt. Sinai Medical Center

John W. Olney, M.D.
Department of Psychiatry and Neuropathology
Washington University School of Medicine

Paul H. Patterson, Ph.D.
216-76 Biology Division
California Institute of Technology

Donald L. Price, M.D.
Neuropathology Laboratory
The Johns Hopkins University School of Medicine

A.H.V. Schapira, D.Sc., M.D., F.R.C.P.
Department of Clinical Neurosciences
Royal Free Hospital School of Medicine

Ira Shoulson, M.D.
Department of Neurology
University of Rochester Medical Center

Caroline M. Tanner, M.D.
Clinical Research
The Parkinson's Institute

Deborah Cordy
Pittsburgh Chapter
American Parkinson Disease Association

Jim Cordy
Pittsburgh Chapter
American Parkinson Disease Association

Morgan Downey, J.D.
National Coalition for Research in Neurological Disorders

Salvatore J. Esposito, Jr.
American Parkinson Disease Association

Lawrence S. Hoffheimer, J.D.
Hoffheimer and Downey

Morton Kondracke

Gianpaolo Maestrone, D.V.M.
Scientific and Medical Affairs
American Parkinson Disease Association

John B. Martin
American Parkinson Disease Association

Peter A. Morabito, D.D.S.
National Parkinson Foundation, Inc.

Dinah Tottenham Orr
Parkinson's Disease Foundation
Columbia-Presbyterian Medical Center

Judy Rosner
United Parkinson Foundation

Lewis P. Rowland, M.D.
Parkinson's Disease Foundation
Columbia-Presbyterian Medical Center
Neurological Institute

Joan I. Samuelson, J.D.
Parkinson's Action Network

Nathan Slewett
National Parkinson Foundation, Inc.

Paul C. Smedberg
Washington DC Regional Office
American Parkinson Disease Association

Norma Gilbert Udall


Workshop Organizers

Thomas N. Chase, M.D.
Experimental Therapeutics Branch
National Institute of Neurological Disorders and Stroke

Carl M. Leventhal, M.D.
Division of Demyelinating, Atrophic, and Dementing Disorders
National Institute of Neurological Disorders and Stroke

Eugene J. Oliver, Ph.D.
Division of Demyelinating, Atrophic, and Dementing Disorders
National Institute of Neurological Disorders and Stroke

Marian Emr
Office of Scientific and Health Reports
National Institute of Neurological Disorders and Stroke

Mary Miers
Science Policy and Analysis Branch
National Institute of Neurological Disorders and Stroke

Curt Pospisil
Science Policy and Analysis Branch
National Institute of Neurological Disorders and Stroke

Additional copies available from

Office of Scientific and Health Reports
National Institute of Neurological Disorders and Stroke
National Institutes of Health
Bethesda, MD 20892


Last updated April 19, 2011