Parkinson Disease: The Ups and Downs of Developing Therapies

April 9, 2005

Levodopa (l-dopa) and dopamine agonists are the main treatment for Parkinson disease (PD), but these therapies are of limited value in the long course of the illness because they counter a neurotransmitter deficit but do not halt neurodegeneration. In this main article and in an accompanying story, we offer an update on the status of Parkinson treatments today.

Levodopa (l-dopa) and dopamine agonists are the main treatment for Parkinson disease (PD), but these therapies are of limited value in the long course of the illness because they counter a neurotransmitter deficit but do not halt neurodegeneration. In this main article and in an accompanying story, we offer an update on the status of Parkinson treatments today.In the 1990 film Awakenings, young neurologist Malcolm Sayer (Robin Williams) treats a memorable group of statuelike patients with l-dopa. The patients return to near normalcy for a short while and then reenter their mysterious catatonic states. The heartbreak of the film is that its story is true, written by neurologist-essayist Oliver Sacks, BM, BCh, and based on a 1973 British documentary of the same name.The patients in Awakenings, the last of whom died in 1992, suffered from encephalitis lethargica, which affected millions in the United States and Europe from 1916 to 1926 and was ultimately traced to a streptococcal infection. When Sacks toured Beth Abraham Hospital in the Bronx, NY, early in his career and saw some of these patients, he was reminded of PD. So when l-dopa began to be used for treatment of PD, he decided to try it on people with the mysterious "sleeping sickness." It worked for a time.Neuroprotective or even neurorestorative therapies remain elusive; but in the past decade, a glimmer of hope has come from those utilizing glial cell line-derived neurotrophic factor (GDNF) therapy. "There is a wealth of preclinical data that suggests that the protein is neuroprotective. This is why we believe it may alter the natural history of the disease, but that has not been established," said Howard Federoff, MD, PhD, director of the Center for Aging and Developmental Biology and professor of neurology at the University of Rochester, New York. You would want to treat PD while it's in a very early stage either to restore function to the residual mass in the substantia nigra in its production of dopamine-secreting neurons or to provoke regeneration in the striatum, said Federoff. "Patients would have to be convinced early in the [course of] disease, when they are still responsive to levodopa, that they should have another form of therapy with a known but small chance of significant complications," he added.Until recently, recombinant GDNF was given-with some success-to patients with late-stage PD; it was delivered directly to the brain using implanted pumps and catheters. Then last September, the manufacturer, Amgen Inc, suddenly put the brakes on phase 2 clinical trials of GDNF, citing serious safety concerns (cerebellar deterioration in monkeys, antibody production in patients) and lack of demonstrated efficacy. Instantly, without one trial in progress, 48 patients could no longer receive the treatment that many claim has helped them-even on a compassionate-use basis and even if they signed waivers absolving Amgen of liability. After intense investigation, Amgen reiterated its stance on February 11. The patients, however, have organized, with many of their physicians behind them. Their stories are posted at Svendsen, PhD, director of stem cell research at the University of Wisconsin-Madison, part of the team that conducted one of the trials, expresses a view shared by many. "The safety data are worrying and serious enough to prevent new surgeries and pump installation. However, it should not limit the current patients from choosing whether to continue GDNF treatment. Some of our patients have been on GDNF for 3 years with no side effects."In the current climate in which the adverse effects of some drugs emerge years after they hit the market, Amgen is using caution. "We don't want to wait for irreversible brain damage to happen. It wouldn't be responsible to lower the standard of a clinical trial on behalf of these patients. It was a challenging decision. It didn't make anyone feel good. I don't mean to minimize what the patients are going through, but clearly this is the right thing to do," Amgen spokesperson Andrea Rothschild told Applied Neurology.The GDNF story is complex, and nowhere near resolution. It began just when another chapter in the search for a PD cure closed.THE AGE OF FETAL TRANSPLANTSThe current controversy over the use of human embryonic stem cells in regenerative medicine was presaged by a more disturbing potential therapy tried a decade ago-use of implants from human late-stage embryos and early fetuses to treat PD. A multicenter team that included Curt Freed, MD, now director of the Neurotransplantation Program for Parkinson's Disease at the University of Colorado School of Medicine, and Stanley Fahn, MD, chief of the Movement Disorders Division and H. Houston Merritt Professor of Neurology at Columbia University, implanted mesencephalon material from 6- to 10-week-old embryos and fetuses in 20 patients, while 20 control patients underwent sham surgery.1At 1 year, the transplant group fared slightly better than the control group: 0 plus/minus 2.1 on the Unified Parkinson's Disease Rating Scale (UPDRS), compared with 20.4 plus/minus 1.7 (negative indicates worsening, positive indicates improvement). For some treated patients under age 60, rigidity and bradykinesia improved but tremor and abnormal gait persisted. Improvements were seen in the morning, before medications were taken. Positron emission tomography (PET) scans using 18F-fluorodopa revealed outgrowths of striatal neurons, indicating that the transplants had worked, but results recorded in the second year suggested that they may have worked too well-or worked in the wrong places. Irreversible dystonia and dyskinesia developed in 5 of the 20 treated patients. In an editorial accompanying the report, Gerald Fischbach, MD, and Guy McKhann, MD, of the National Institute of Neurological Disorders and Stroke cautioned that such trials must go on long enough to rule out delayed responses.2Despite some success in European trials, fetal transplant research quietly vanished and in its place, at least coincidentally, arose a search for the molecules in the transplanted tissue that stimulated survival of striatal neurons and dopamine secretion. The warning to watch for delayed effects, however, would reverberate.GDNF ARRIVESIn the 1980s, following the ever-growing list of newly identified growth factors, came interest in neurotrophic factors. Unlike the growth factors that stimulate cell division, neurotrophic factors nurture selected cell types, keeping them alive and active. Toward the end of the decade, Yves-Alain Barde, PhD, and coworkers3,4 at the University of Basel, Switzerland, discovered brain-derived neurotrophic factor (BDNF), which stimulates neurite growth and dopamine secretion.In 1993, about the same time that the patent for BDNF was issued, Frank Collins, PhD, Leu-Fen Lin, PhD, and colleagues5 at Synergen Corporation in Boulder, Colo, were searching for neurotrophins in glia, specifically B49 rat glioblastoma cells. It took biochemist Lin 6 months of meticulous work to isolate sufficient GDNF to decipher enough of the amino acid sequence to synthesize a DNA probe. The researchers then pulled out the GDNF gene from a human gene library and scaled up protein production in recombinant bacteria. Biochemist Jack Lile coaxed the protein to fold properly, and other experiments showed that it allowed selected survival of dopaminergic neurons beyond the normal 3-week life span in culture. GDNF also quickened maturation of neurons in embryos and increased sprouting in just days, and at low concentrations. The last sentence of the research paper by Collins, Lin, and colleagues, which was published in Science in 1993, suggested that GDNF may be useful in the treatment of PD.At the end of 1994, Amgen acquired Synergen. Less than a month later, researchers affiliated with the company reported the ability of GDNF to promote survival of cultured human fetal dopaminergic neurons.6 By May 1995, patent filings began that claimed a variety of uses for GDNF and addressed the inability of GDNF both to cross the blood-brain barrier and to be taken orally. Direct delivery to the brain seemed to be the way to go, as Swedish researchers had done for nerve growth factor to treat PD and Alzheimer disease. A year later, an investigation headed by Don M. Gash, PhD,7 chair of Anatomy and Neurobiology at the University of Kentucky, demonstrated the efficacy of GDNF in a monkey model of PD.With this encouraging track record, GDNF wound its way into clinical trials. In 1999, Amgen researchers tried delivering the protein to the ventricles in 2 successive groups, each consisting of 30 patients with PD. No improvement was seen, but many adverse effects were-possibly caused by the brain becoming awash with the protein. The endeavor clearly indicated that direct delivery to the affected part of the brain was key.An off-label trial in 5 patients began in 2001 under the direction of consultant neurosurgeon Steven S. Gill, FRCS, at the University of Bristol Institute of Clinical Neuroscience at Frenchay Hospital in the United Kingdom. Catheters were inserted into the putamen through 2 drilled holes, and GDNF was delivered from pumps implanted in the abdomen and threaded upward. All 5 patients in the phase 1 safety trial improved, some dramatically.8 One participant, Stephen Waite, had great difficulty in rising from a chair and propelling himself forward before treatment; after treatment, he could walk.The success of the Bristol University trial inspired another open-label trial of GDNF, this one at the University of Kentucky. The dose was increased every 8 weeks, with intervening 5-week washout periods. At the 6-month mark, "off" and "on" scores on the UPDRS had improved by 34% and 33%, respectively, for the 10 patients participating in the study.9 The effect was sustained through washout, and only 2 participants experienced side effects (transient Lhermitte sign). PET scans indicated a slowing of the loss of dopaminergic neurons. Patients wrote to Amgen chronicling their improvement. One participant, a teacher, described how she could once again fold towels, button buttons, and smile at strangers.The University of Kentucky trial, however, was not placebo-controlled. So, in 2003, Amgen initiated a multicenter placebo-controlled trial that included 34 patients, all of whom received the pump-and-catheter delivery system. For the first 6 months, half the group received saline solution; then, for ethical reasons, they were switched to GDNF. At the end of 6 months, in June 2004, not only were the improvements not statistically significant, according to the company's analysis, but some of the patients in whom the best results had been seen were from the placebo group. Since the drug seemed safe at that time (August 2004), Amgen considered increasing the dose, according to Rothschild.Then all hell broke loose.THE CONTROVERSY BEGINSThe safety issues that arose in August 2004 were indeed serious. "A number of patients had neutralizing antibodies, and we also found in toxicology studies irreversible brain damage in the preclinical model, using a dose that was very close to the dose that had already been used in open-label trials and the dose we were considering using in patients," said Rothschild. Of the 70 monkeys, 5 in the group of 15 given the highest dose experienced cerebellar degeneration, she said.EXHAUSTIVE EVALUATIONThroughout the fall of 2004, Amgen analyzed and reanalyzed the toxicity data and met with investigators inside and outside the company. "We called in bioethicists and asked, 'if we provide this drug to patients, will we be waiting to see what we saw in preclinicals?' Considering the lack of benefit in phase 2 [trials] and 2 significant safety findings, we felt we couldn't continue with human toxicity studies," Rothschild said. The ghost of the fetal tissue experiments of a decade earlier loomed. "It was the same invasive surgery, the same hope of changing the underlying disease rather than treating symptoms. At first the fetal cells seemed to dramatically improve symptoms, so sham surgeries were necessary to debunk the assumptions. There was delayed toxicity," she explained.While all parties concur that the warning signs are cause for concern, the interpretations are far from clear. Fuzziness of detail, possible trial design differences, and the willingness of patients to accept risks have spawned an emotionally charged battle. At the heart of the issue is the significance of the safety matters.A consensus statement from investigators at clinical trial centers (the University of Kentucky, Frenchay Hospital, New York University, and the University of Chicago) counters Amgen's perspective: "Cerebellar toxicity in rhesus monkeys occurred outside of the clinically relevant dose range. Antibody expression frequently occurs with other proteins used to treat disease of the brain without producing recognizable clinical autoimmune disease and without precluding clinical treatment. First and foremost, to date none of the patients have shown evidence of cerebellar toxicity or autoimmune symptomatology." The group advised close monitoring of patients and further testing-not halting-their treatment. Reported facts vary. The New York Times quoted Amgen as citing brain damage in 4 monkeys, and various groups place the monkeys' dose at 12 to 15 times the human dose. Some researchers have suggested that the cerebellar degeneration might have been caused by abrupt treatment withdrawal in the study animals.Another point of contention is consistency of experimental design. "Our initial trial used a fine catheter that caused little damage to the brain and achieved convection-enhanced delivery of GDNF.8 The Amgen trial used a much larger catheter and may not have achieved wide spread of the drug. They also used lower doses of GDNF and a different patient group," said Svendsen. The consensus statement echoes his opinion that infusion procedures using convection-enhanced delivery are essential: "The methods used in the two phase 1 studies achieved this goal, with 15 of 15 advanced Parkinson's disease patients showing significant functional improvements. The phase 2 study did not replicate the phase 1 studies in these 2 parameters of dose and tissue distribution."Kristen Suthers, a statistician for the Johns Hopkins University Phoebe R. Berman Bioethics Institute, whose father is 1 of the 48 patients, questioned the data analysis. "They used insufficiently longitudinal data on a normal distribution-but that's not the appropriate distribution," she said.Some patients' families mention the matter of finances. Furthermore, such an invasive treatment might not have widespread use and research is already several years into patent life. However, the consensus statement maintains that orphan drug status applies to a subset of about 200,000 patients in the United States who are nondemented and have advanced, life-threatening disease. Also, Amgen has already invested many millions of dollars in developing GDNF.Amgen officially disagrees with the criticism of the trials' design and results. "There's debate about the phase 2 trial: the dose was not high enough, the catheters different. But the bottom line is, it was a well-controlled trial designed to replicate one of the open-label studies," explained Rothschild. She added that to find neutralizing antibodies is reason enough to end a trial or pull a marketed drug. The company also denies that financial motivation played any role in discontinuing the trial and supplying the drug on a compassionate-use basis, and it refutes the suggestion that the cyclooxygenase-2 inhibitor controversy fueled recent decisions. "Those companies are being questioned for their ethics around promoting drugs in the face of compelling clinical evidence and concerns. We are doing the opposite," she added. The discontinuation of the GDNF trial preceded Pfizer's Vioxx woes of September 2004.Despite the ongoing disagreements and disappointments, researchers and patients concur that more study is needed and that GDNF still holds great promise. Indeed, Amgen is among those continuing to support investigation of new ways to deliver the protein. However, GDNF trial coordinators maintain that efforts to use other approaches are 5 to 10 years from the clinic. Part of the challenge is to design new, appropriate trials. "Can you recruit subjects at an early stage and in significant enough numbers to conduct a phase 2 or 3 trial, assuming that safety and tolerability have already been shown in trials on advanced patients?" asked Federoff.Still, against the backdrop of lessons learned from fetal transplant experiments and the recent GDNF safety concerns, the explosion of new information on the molecular profiling of stem cells may finally make the time ripe for effectively utilizing GDNF. Transplanting fetal mesencephalon tissue may have introduced a mix of cell types, but delivering pure neural stem cells, appropriately genetically modified to create GDNF-secreting astrocytes, may more closely mimic the natural trajectory and distribution of this protein in the healthy human brain. Teaming stem cell biology with genetic modification harnesses the controls of GDNF delivery-likely the source of continuing problems in the PD drug arsenal. Svendsen's group is trying this approach on animal models of PD, discussed in the January issue of Applied Neurology.Ron Mandel, PhD, at the Powell Gene Therapy Center and the McKnight Brain Institute at the University of Florida in Gainesville is pursuing a gene therapy approach, with collaborators from Lund University in Sweden and Cambridge University in the United Kingdom. In work to be reported in the Journal of Neuroscience, they delivered GDNF to the striatum using an adeno-associated virus as a vector. "It saves the innervation of the neostriatum in rodent and primate Parkinson's models," Mandel said. Treated monkeys improved at various tasks, and brain scans indicated prevention of cell death in the substantia nigra. "Our system is different from Amgen's in that we use gene transfer to make a continuous level of GDNF available, at much lower concentrations and only to the right part of the brain," he said. The effect appears to be neuroprotective, but it will not be ready for clinical trials until the researchers can learn how to turn it off, he added. It might one day be useful in patients with early-stage disease.Researchers from the Department of Neurological Surgery at Okayama University in Japan found a similar neuroprotective effect when they introduced encapsulated human GDNF-secreting cells in hollow fibers into brains of adult rats given a drug to induce parkinsonism. The rats treated earliest showed the most dopamine secretion and glial proliferation in the striatum.10The GDNF saga is not over. The families and their physicians are continuing discussions with the FDA and are seeking a new sponsor, although Amgen does not want to go this route for fear of liability for long-term adverse effects. The consensus statement sums up the situation by calling GDNF delivered by the pump-and-catheter method "the bird in the hand." But whether that bird will fly remains to be seen. Said Federoff, "We can't yet know if GDNF will supplant or be an adjunct to conventional therapy. We can only speculate on whether it would flatten or slow disease progression."William Weiner, MD, a neurologist and director of the Maryland Parkinson's Disease and Movement Disorder Center in Baltimore, raised longer-term concerns. "Let's say GDNF works, and saves dopamine-secreting cells. What does that mean? That motor problems will not get bad as quickly? That tremors won't happen? Is GDNF as a specific therapy that saves dopamine cells really enough? It is painfully obvious to everyone who treats patients with Parkinson disease that it is not an isolated dopamine problem. Much more widespread neurodegeneration eventually takes over, and this is why people become unresponsive to levodopa. It is why they become demented, [and why they] develop hypotension and erectile dysfunction. If we can save the dopamine-secreting cells, what will that mean 10 to 15 years out?"Ricki Lewis, PhD, a geneticist, textbook author, and freelance writer in Scotia, NY, is a contributing editor for Applied Neurology.REFERENCES1. Freed CR, Greene PE, Breeze RE, et al. Transplantation of embryonic dopamine neurons for severe Parkinson's disease. N Engl J Med. 2001; 344:710-719.2. Fischbach GD, McKhann GM. Cell therapy for Parkinson's disease. N Engl J Med. 2001;344:763-765.3. Hofer MM, Barde YA. Brain-derived neurotrophic factor prevents neuronal death in vivo. Nature. 1988;331:261-262.4. Leibrock J, Lottspeich F, Hohn A, et al. Molecular cloning and expression of brain-derived neurotrophic factor. Nature. 1989;341:149-152.5. Lin LF, Doherty DH, Lile JD, et al. GDNF: a glial cell line-derived neurotrophic factor for midbrain dopaminergic neurons. Science. 1993; 260:1130-1132.6. Tomac A, Lindqvist E, Lin EF, et al. Protection and repair of the nigrostriatal dopaminergic system by GDNF in vivo. Nature. 1995;373:335-339.7. Gash DM, Zhang Z, Ovadia A, et al. Functional recovery in parkinsonian monkeys treated with GDNF. Nature. 1996;380:252-255.8. Gill SS, Patel NK, Hotton GR, et al. Direct brain infusion of glial cell line-derived neurotrophic factor in Parkinson disease. Nat Med. 2003;9:589-595.9. Slevin JT, Gerhardt GA, Smith CD, et al. Improvement of bilateral motor functions in patients with Parkinson disease through the unilateral intraputaminal infusion of glial cell line-derived neurotrophic factor. J Neurosurg. 2005;102:216-222.10. Yasuhara T, Shingo T, Muraoka K, et al. Early transplantation of an encapsulated glial cell line-derived neurotrophic factor-producing cell demonstrating strong neuroprotective effects in a rat model of Parkinson disease. J Neurosurg. 2005;102:80-89.---Sidebar-A Personal View of the Impact of the GDNF Trial CessationViews of the participants in the phase 1 and 2 clinical trials for glial cell line-derived neurotrophic factor (GDNF) and their relatives, as well as letters of support from investigators and organizations, can be accessed at the Web site These testimonials complement the scientific data. Kristen Suthers, a statistician at the Johns Hopkins University, Phoebe R. Berman Bioethics Institute, speaks about her father Robert's experience:"My father's symptoms began in 1997, and he was diagnosed and began treatment in the fall of 1998. He took the standard levodopa and was constantly changing drugs, trying different things. Then in the summer of 2002, he started to decline faster. He had been stable for 4 to 5 years on regular [Parkinson disease] meds, but the drugs stopped working. He had said he'd never undergo brain surgery, but once the symptoms worsened, he decided to go ahead. The cutoff age for the study was 70, and he was 69. It was now or never. In spite of the stroke he had from the procedure, he never regretted participating in the trial-until they pulled the plug in early September. Then they added saline to his pump to get the last little bits.He's been off GDNF since September, and we have watched his slow decline back to where he was before. Yes, it is like Awakenings. It is horrific to watch someone you love go through this. I don't know if my dad will ever get the drug. He's slipping fast. But there are people in their 40s who have Parkinson's. They are trapped in their bodies."---Sidebar-New Drug in Arsenal for Parkinson Disease on the Near HorizonResults of the final arm of the Parkinson's Rasagiline: Efficacy and Safety in the Treatment of "Off" (PRESTO) trial indicate that rasagiline (Agilect, Teva Neuroscience) may be useful in a range of patient populations affected by Parkinson disease. The drug is a novel, irreversible, selective monoamine oxidase type B (MAO-B) inhibitor.The recent study complements a previous trial of 404 patients that found that the drug was well tolerated and more effective than placebo.1 "First we published results on patients who had early disease and didn't require any dopaminergic therapy. This report follows a separate group of patients who are more advanced and are on levodopa (l-dopa) or dopamine agonists.2 We are interested in ensuring that this drug, which is safe and effective in an [early-stage disease] population, is likewise safe and effective in the more [advanced-stage] population. They are more vulnerable to adverse effects due to potential drug interactions," explained Ira Shoulson, MD, a professor in the Department of Neurology at the University of Rochester Medical Center in New York and principal investigator of the Parkinson Study Group that conducted PRESTO.This latest investigation involved 472 patients at 57 study sites in the United States and Canada and lasted from December 2000 to June 2002. Participants were patients who experienced at least 2.5 hours a day of "off" periods, when medications are ineffective in controlling symptoms. The study compared high-dose rasagiline (1 mg/d), low-dose rasagiline (0.5 mg/d), and placebo.The drug improved motor fluctuations and symptoms. "Off"-period duration fell by 1.85 hours daily in the high-dose group, 1.41 hours in the low-dose group, and 0.91 hour in the placebo group. Improvements also appeared on 2 rating scales: the physician-delivered clinical global impression scale and a patient-completed activities of daily living scale.Rasagiline has an interesting history. Moussa Youdim, PhD, professor of life sciences at Technion, the Israel Institute of Technology, developed it to treat hypertension and depression at the University of London in 1969. Although the agent was not effective for these indications, Youdim thought the drug might have anti-Parkinson potential and pursued that avenue of research in Israel. But in 1977, when he approached Teva Neuroscience, the small generic drug company was not ready to test the compound. A decade later, the company had grown and was in search of projects, and Youdim got his foot in the door.As an MAO-B inhibitor, the drug blocks some metabolism of dopamine and potentiates the activity of l-dopa. Yet, it also shows neuroprotective and neurorescue activities in vitro and in vivo, said Youdim. "The neuroprotective and neurorescue activity is not related to its MAO inhibitory activity, since the optical isomer of rasagiline, which is a very poor inhibitor, has similar neuroprotective activity. In studies where rasagiline was chronically given after MPTP [1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine] caused degeneration of dopamine neurons, rasagiline caused neurorescue of the neurons. This is attributed to its ability to induce GDNF [glial cell line-derived neurotrophic factor] and BDNF [brain-derived neurotrophic factor] mRNA and activate the Akt pathway," he added.Rasagiline is just about to reach the market in Israel and Europe. In the United States, the FDA has sent a letter of approval to Teva Neuroscience, and clearance for marketing is anticipated for later in 2005. The drug is taken as a daily tablet.REFERENCES1. Parkinson Study Group. A controlled, randomized, delayed-start study of rasagiline in early Parkinson disease. Arch Neurol. 2004;61:561-566.2. Parkinson Study Group. A randomized placebo-controlled trial of rasagiline in levodopa-treated patients with Parkinson disease and motor fluctuations: the PRESTO study. Arch Neurol. 2005;62:241-248.--------------------------------------------------------------------------------