Seeking Biomarkers for Parkinson’s

by Carl Sherman

November 14, 2013

Although treatments for people with Parkinson's disease (PD) has improved in recent years, no drug has emerged that can alter the relentless course of the second-most-common neurological disorder, let alone prevent it.

Among the obstacles: the affected neurons start to die apparently undetected for years, when PD may be most treatable; once symptoms emerge, they evolve slowly, making it hard to test drugs; and misdiagnosis can confuse research findings (an estimated 15 percent of patients in clinical trials actually have another motor disorder).

The identification of anatomical and physiological manifestations-biomarkers-that appear earlier and change more quickly than symptoms would vastly improve our ability to track the course of PD and advance drug development and clinical care, says Katrina Gwinn of the National Institute for Neurological Disorders and Stroke.

"Research in this area is more active than I've ever seen it in the past," she says, thanks in part to progress in fields like proteomics, which allows more precise analysis of body fluids [like serum and cerebrospinal fluid (CSF)], and neuroimaging.

Recent Alzheimer's research, particularly the Alzheimer's Disease Neuroimaging Initiative (ADNI), that have resulted in biomarkers already in use, has provided inspiration and a model, Gwinn says.


A landmark trial

The Michael J. Fox Foundation's Parkinson's Progression Markers Initiative ( PPMI) is a large, international observational trial that will track recently diagnosed PD patients over five years.

"Our primary goal is to help develop drugs to slow, stop, or even prevent PD," says Kenneth Marek of the Neurodegenerative Disease Institute in New Haven, and principal investigator of PPMI. "Since we recognize that PD is likely a number of different if related problems that lead to a single entity, a secondary goal is to identify biomarker signatures that could lead to specific treatments for these subtypes."

 PPMI's 24 participating centers have completed baseline clinical and biological data collection for 400 recently diagnosed patients and 200 healthy controls. Follow-up blood and CSF specimens and neuroimaging data will be taken every three and then six months for five years.

"We're hoping to validate biomarkers that already have a modicum of data," says Marek-some 20 to 25 thus far. 

The first analysis of data from PPMI, a study of CSF samples in the first 63 patients and 39 controls enrolled, appeared in the Aug. 26, 2013, issue of JAMA Neurology.

"The goal of this pilot study was to see if gathering CSF samples [a somewhat arduous procedure] would be worth the trouble," says Marek. "It was surprising to find that even with a relatively small sample size and few assays, we could detect a biomarker signature."

The study looked at alpha-synuclein, a component of Lewy bodies, the distinctive pathological structure in PD; tau protein, a signature protein in PD as well as Alzheimer's disease; and amyloid beta, the building block of Alzheimer's plaques.

"Some 80 percent of PD patients will develop dementia… due to progressive evolution of Lewy bodies into cognitive areas or, in one-third of cases, to Alzheimer's pathology," says John Trojanowski, director of the University of Pennsylvania's Udall Center for Parkinson's Research and a senior author of the paper. "So it's essential to monitor these proteins."

"For PD disease-modifying treatment research, we desperately need biomarkers that reflect core pathologies. It would be important to know, for example, if we ameliorate alpha-synuclein, whether plaques and tangles would respond as well," says Trojanowski, who is also a member of the Dana Alliance for Brain Initiatives.

In the sample, concentrations of all four proteins were significantly different in the CSF of PD patients and controls, most strikingly in those patients in whom postural instability and gait disturbance predominated-a subtype more likely to develop dementia.

"What made this paper important was that it showed that each of these biomarkers changed at such an early stage in PD-these patients had clinical and neuroimaging evidence of the disease, but it wasn't severe enough to require medication," Trojanowski said. By following CSF data on all 600 patients and controls for five years, the team hopes to see if such early changes may predict later cognitive impairment and conversion to dementia.

 "If we want drugs that will have a great impact on the disease, we think that will come when the disease is mild, or even in a prodromal stage," Trojanowski said.

In fact, a new arm of PPMI has recently started recruiting people who are at high risk of the disease but do not yet have its distinctive motor symptoms, says Marek. Because degeneration in the brain begins well in advance of symptoms, "the best group to treat would be those who don't yet have the disease," if biomarkers could identify them.


A panoply of projects

At the end of 2012, the National Institute on Neurological Disorders and Stroke announced its own initiative, the Parkinson's Disease Biomarkers Program, to explore new ways to track the disorder. "The focus of our project is to develop biomarkers for clinical trials [that will lead to] neuroprotective agents," says Katrina Gwinn, lead program director.

"We didn't want to reproduce the Fox Foundation work: They're much closer to the endgame, trying to get FDA approval of biomarkers for use in trials and the clinic. It looked to us that more discoveries were needed to go into the pipeline," she says. The PDBP is currently funding 10 independent projects covering a spectrum of research.

Alice Chen-Plotkin, of University of Pennsylvania, is screening hundreds of proteins in search of those that might correlate with clinical manifestations of PD. "The whole concept of investigation on this scale only became possible in the mid-'90s," originating in genomic research, she says.

Her project in the program includes further exploration of Apoprotein A1, identified as a possible biomarker by prior large scale screening. In the July 2013 issue of Annals of Neurology, Chen-Plotkin and colleagues reported associations between lower levels of ApoA1 and earlier age of onset in people with PD, and with dopaminergic system vulnerability in individuals at increased PD risk.

Other lines of research, she points out, have implied a role for the protein in PD pathophysiology: e.g. cardiovascular studies indicate that statins, which increase Apo A1 (a key component of HDL-cholesterol), reduce incidence of the movement disorder. "What's so neat about ApoA1: if we see that it's not only a marker for PD but on the causal chain, we already have drugs that can alter levels," she says.

Her NINDS-funded research will look at ApoA1 levels in much larger populations, and perhaps investigate mouse models of PD in animals genetically modified to over- or underexpress the protein.

In his PDBP project, David E. Vaillancourt, of University of Florida, is investigating the utility of neuroimaging to identify PD and track its progression.

These studies build on his earlier research; in one series of studies, reported in Movement Disorders in May 2013, Vaillancourt and colleagues showed that diffusion tensor imaging patterns could  differentiate patients with PD from those with other movement disorders and healthy controls. A 2013 paper in JAMA Neurology identified fMRI patterns associated with PD subtypes, including a variant that typically progresses rapidly with a high likelihood of dementia.

He and colleagues are attempting to replicate these findings with more patients, and following some a year later to see if and how functional and structural changes correspond to clinical patterns. "A goal is to understand how the disease spreads from one area of the brain to another… and to develop markers for these changes with which to test interventions," Vaillancourt says.

Other PDBP projects include postmortem tissue studies, DNA analysis, and computational dissection of a huge database of medical records.

More than any single parameter, Gwynn says, it is likely that "a cocktail of biomarkers" will together provide a powerful tool for tracking PD.

"We designed the program to build a resource as well as ask research questions," she says. Specimens and data from NIMH projects will be made available to the wider PD research community, whether or not they are associated with the program. A principal aim of the Fox Foundation's PPMI is, similarly, to vitalize biomarker validation by providing a large databank that researchers can use to test and explore findings from an array of studies.

"We want to encourage investigators to go to our website and sink their teeth into our data," says Marek.