Read a PDF of the entire series.
27 March 2013. Adding insult to injury, many patients with a movement disorder get dementia, too. In fact, up to 80 percent of Parkinson’s disease (PD) patients develop cognitive deficits (see Aarsland et al., 2003). But researchers do not know what causes this mental decline. Is it α-synuclein pathology in the cerebral cortex? Aβ pathology from comorbid AD? Dying dopamine cells in the basal ganglia? Or all of the above? At the 11th International Conference on Alzheimer’s and Parkinson’s Diseases, held 6-10 March 2013 in Florence, Italy, several groups showed how they are using imaging and biomarkers to parse the causes of dementia in Lewy body disorders. The more tools scientists have to look, the more the answer appears to be: all of the above.
The topic is picking up in the scientific literature as well. On March 25 in JAMA Neurology, scientists reported that mild cognitive impairment (MCI) in early PD predicts dementia down the road. Other researchers reported in the journal Movement Disorders that cerebrospinal fluid (CSF) Aβ in PD correlates with enlarged brain ventricles.
For clues about whether Aβ causes dementia in Lewy body disorders, John Growdon, Massachusetts General Hospital, Boston, told the audience at AD/PD 2013 that his group conducted cognitive tests and amyloid imaging with Pittsburgh Compound B (PIB) on normal aged controls and patients with PD, PDD, or DLB. PDD is Parkinson’s disease dementia, and DLB, dementia with Lewy bodies. Groups comprised 12 to 29 people. The most fibrillar amyloid occurred in the brains of people with DLB, followed by PDD and PD without dementia. In the DLB group, having more amyloid correlated with a greater clinical dementia rating and lower MMSE score. Those results suggest that in DLB, amyloid pathology is one of the abnormalities leading to dementia, Growdon said.
In this case-control, one-time comparison, amyloid pathology did not distinguish PD patients with mild cognitive impairment from those without. However, PET imaging with 18F-fluorodeoxyglucose did. (FDG) PET picks up hypometabolism. This supports the notion that both groups may have amyloid early on, but only those who also have synapse loss in the cerebral cortex have developed MCI, Growdon said.
Does having amyloid pathology when one has Parkinson’s predict that cognitive decline will follow? To find out, Growdon and colleagues conducted baseline amyloid scans in 35 cognitively normal people with PD and 11 people with PD-MCI, and administered cognitive tests every year for two and a half to five years. While baseline PIB uptake still did not distinguish between PD and PD-MCI groups, it did predict the speed of progression. Those who had PIB uptake above the median, i.e., had more amyloid in their brains, progressed faster to PD-MCI or PDD than those with uptake below the median (see Gomperts et al., 2013). “These results establish β amyloid as a risk factor for causing dementia in PD,” Growdon told listeners in Florence. “Whenever effective anti-amyloid therapies emerge from research for AD, such treatments will be immediately relevant for PD as well.”
Amyloid-β pathology is not the only contributor to cognitive impairment in people with Parkinson’s. Dopaminergic cell death spreading to areas important for learning and cognition also plays a part. Growdon used the PET radioligand altropane, which labels dopamine transporters, to find out if a loss of this synaptic signal was linked with worse cognition. In both DLB and PDD patients, a lower signal in the caudate correlated with cognitive impairment, suggesting that a loss of dopaminergic neurotransmission in that structure contributes to dementia in DLB and PDD, Growdon said.
Studies like this are important for designing clinical trials, said John Trojanowski, University of Pennsylvania, Philadelphia. “We have a spectrum of clinical manifestations of Lewy body pathology that can be PD, PDD, or things in between,” he told Alzforum. “Yet the emergence of dementia in these patients is not due just to Lewy bodies; there are other contributors, like plaques and tangles.” To target patients at risk for dementia in clinical trials, researchers need to understand which pathologies predispose to cognitive decline in PD, he said.
Can imaging aid in the differential diagnosis of these disorders? To find out, Andrew Siderowf, who left the University of Pennsylvania, Philadelphia, to work for Avid Radiopharmaceuticals, Philadelphia, imaged five PD patients, 11 people with DLB, 10 with AD, and five healthy controls with the amyloid tracer 18F-florbetapir and with AV-133, a PET ligand for vesicular monoamine transporter 2 (VMAT2). AV-133 is an investigational tracer. It images the protein that loads synaptic vesicles with monoamine neurotransmitters such as dopamine (see ARF related news story). Avid, which developed florbetapir (trade name Amyvid®), is developing AV-133 for clinical use in differential diagnosis. In this small study, patients with PD and DLB had abnormal AV-133 scans, much as might be expected with altropane, while those with AD and healthy controls had normal ones. “AV-133 imaging may be useful for ruling out AD because, in general, patients with AD have normal scans,” said Siderowf.
Most clinically diagnosed AD patients had evidence of amyloid deposition. DLB patients either had both amyloid and dopaminergic loss, or only the latter, making them a mixed group. In addition, patients with less dopamine signal in the caudate had a lower MMSE score. “That could be a biomarker signal that explains part of why patients with DLB have cognitive problems,” Siderowf told Alzforum, noting that his results fit well with Growdon’s. “The association between cognition and dopamine loss in the caudate in Lewy body disorders has not been widely reported,” he said. “It’s reassuring to know that our results are consistent.”
Amyloid and dopamine imaging in people with DLB may help shed light on the heterogeneity of this disorder, and aid prognosis and treatment in the future, he added. In this early imaging study, the participants’ diagnoses were not autopsy confirmed.
Both research groups necessarily compared apples and oranges. For AD, they had a way to visualize a defining protein pathology but not cholinergic neurotransmission, whereas in Lewy body disorders, they are able to image damage to dopaminergic neurotransmission but not the defining protein pathology. Imaging agents to visualize tau, a defining pathology in AD that is also thought to be involved in some Lewy body disorders, are only just entering human testing.
In the absence of a complete set of imaging tools, levels of cerebrospinal fluid (CSF) Aβ42 may also predict cognitive decline in PD, according to Lucia Farotti, who works with Lucilla Parnetti at the University of Perugia, Italy. Farotti monitored CSF biomarkers and cognition longitudinally in 56 patients with PD. If participants' Aβ42 exceeded 800 picograms/milliliter in the CSF at baseline, their MMSE scores stayed stable up to eight years. Those with levels under 800 pg/ml at baseline declined by an average of two points during that time. CSF Aβ42 levels predicted cognitive decline in PD patients with a sensitivity of 80 percent and a specificity of 54 percent, Farotti said. These results add to the argument that AD pathology is a predictor of cognitive impairment in PD, and could be a useful prognostic biomarker, she concluded. The data confirm previous findings by other labs of lower CSF Aβ in some people with PD, and add longitudinal observation.
Seen clinically, mild cognitive impairment, too, predicts dementia in PD. That is according to one of the first studies on the prognosis of MCI in PD, published online March 25 in JAMA Neurology. A previous report on advanced PD patients found that 62 percent of those with MCI progressed to dementia over four years, versus 20 percent who did not have MCI at baseline (see Janvin et al., 2006). This is the first study to examine the same issue in early PD, said Kenn Freddy Pedersen, Stavanger University Hospital, Norway. He and colleagues reported three-year data from the prospective Norwegian ParkWest study, which followed 182 non-demented patients with newly diagnosed PD for three years. By the third year of follow-up, 27 percent of those with MCI at baseline were diagnosed with dementia, compared to only 0.7 percent of those without. Translating to a 9 percent annual rate of progression, this estimate roughly equals that found for AD (see Petersen et al., 2009), wrote the authors.
Some participants reversed their cognitive decline. One in five people with PD-MCI at baseline tested normally at the three-year follow-up, again within limits of the 14-41 percent range reported for non-PD populations. Of those who had PD-MCI at both baseline and one year, 45 percent progressed to dementia after three years and only 9 percent reverted. That suggests repeated tests strengthen the prognostic value of MCI, wrote the study authors. With a sensitivity of 91 percent and a negative predictive value of more than 99 percent, MCI within the first year of PD diagnosis is a strong predictor for further cognitive decline, they concluded.
That many people reverting to normal cognition “suggests that PD-MCI is not a stable finding,” commented Brian Copeland and Mya Schiess of the University of Texas Medical School at Houston, in an accompanying editorial. “Any enthusiasm over the predictive value of the diagnosis must be tempered with that in mind,” they added. MCI is a clinical diagnosis.
Another hint on how to sharpen such a clinical diagnosis came just this month from the same ParkWest study in Norway. In this month’s issue of Movement Disorders, Mona Beyer at Stavanger, working with imaging researchers led by Liana Apostolova at the University of California, Los Angeles, link low CSF Aβ to ventricular enlargement in the study participants. Since expanding ventricles were previously associated with disease progression in MCI, AD, and PD, CSF and MRI markers may help identify patients at risk for cognitive decline, wrote the authors.
In recent years, Tamas Revesz and colleagues at the Queen Square Brain Bank (QSBB), London, U.K., have published data about dementia in PDD gleaned from histopathological studies of postmortem human brains. In an AD/PD 2013 conference presentation reviewing these findings, Revesz highlighted a role for Aβ and tau pathology in addition to Lewy bodies. All three pathologies are present in those who died with PDD, and a combination better predicts dementia than any one of them alone. Interestingly, Aβ and α-synuclein seem to interact, where a higher level of one tends to come with more of the other. “If you have a treatment that modifies α-synuclein pathology, that one therapy may not be sufficient to treat these individuals. You may also need to target Aβ pathology,” said Revesz, echoing Growdon’s statement about the potential value of future anti-amyloid medications for patients at risk of cognitive decline in PD.
These studies are important because they generate data on larger numbers of participants that together enable scientists to understand what is causing dementia in Lewy body disorders, said Trojanowski. “We need more autopsy data with longer and better follow-up of patients within the last few months or years of life to get a more secure idea of what causes the dementia.” With such information, scientists may be able to identify who is going to get PD-related dementia before they exhibit symptoms, said Pedersen, adding, “If we could reach these patients before they become impaired, we could perhaps delay dementia.”—Gwyneth Dickey Zakaib with reporting from Madolyn Bowman Rogers.
Pedersen KF, Larsen JP, Tysnes O, Alves G. Prognosis of Mild Cognitive Impairment in Early Parkinson Disease: The Norwegian ParkWest Study.
JAMA Neurol 2013 March 25. Abstract
Copeland BJ and Schiess MC. Can Mild Cognitive Impairment in Parkinson Disease Predict the Development of Dementia? JAMA Neurol 2013 March 25. Abstract
Beyer MK, Alves G, Hwang KS, Babakchanian S, Bronnick KS, Chou YY, Dalaker TO, Kurz MW, Larsen JP, Somme JH, Thompson PM, Tysnes OB, Apostolova LG. Cerebrospinal fluid Aβ levels correlate with structural brain changes in Parkinson's disease. Mov Disord. 2013 Mar;28(3):302-10. Epub 2013 Feb 13. Abstract