New insights into atypical parkinsonism

Abstract

Purpose of review

Atypical parkinsonian disorders (APDs) comprise a heterogenous group of disorders including multiple system atrophy (MSA), dementia with lewy bodies (DLB), progressive supranuclear palsy (PSP) and corticobasal degeneration (CBD). Based on literature published in 2010 we here review recent advances in the APD field.

Recent findings

Genome-wide association studies have provided robust evidence of increased disease risk conferred by synuclein and tau gene variants in MSA and PSP. Furthermore, advanced imaging tools have been established in the differential diagnosis and as surrogate markers of disease activity in APD patients. Finally, although therapeutic options are still disappointing, translational research into disease-modifying strategies has accelerated with the increasing availability of transgenic animal models, particularly for MSA.

Summary

Remarkable progress has been achieved in the field of APDs and advances in the genetics, molecular biology and neuroimaging of these disorders will continue to facilitate intensified clinical trial activity.

Introduction

Movement disorders experts are increasingly challenged by atypical parkinsonian disorders (APDs) that include α-synucleinopathies (multiple system atrophy (MSA), dementia with Lewy bodies (DLB)) and tauopathies (progressive supranuclear palsy (PSP), and corticobasal degeneration (CBD)). APDs are characterized by relentlessly progressive and levodopa refractory parkinsonism associated with a range of distinctive atypical features [1**]. We here review recent developments in APD research covering preclinical as well as clinical aspects.

Synucleinopathies

α-synuclein is the main component of the inclusion pathology in MSA and DLB classifying these two disorders among other α-synucleinopathies such as PD and pure autonomic failure.

Multiple System Atrophy

While there has been relatively little progress in the clinicopathological understanding of MSA a number of transgenic models have recently generated important insights into glial synuclein-associated neurodegeneration that have stimulated interventional neuroprotection trials in MSA patients [2].

Molecular pathogenesis

The working hypothesis of MSA as a primary oligodendrogliopathy with prominent (oligodendro-) glial cytoplasmic inclusions and secondary neuronal multisystem degeneration [3] has recently been corroborated by the discovery of deficient oligodendroglial GDNF release associated with selective neuronal loss in a transgenic MSA mouse model [4]. Extending previous findings (Song 2007) tubulin polymerization promoting protein (TPPP) also known as p25alpha was shown to specifically accelerate oligodendroglial α-synuclein oligomer formation and to promote α-synuclein-positive GCI-like inclusions [5]. The concept of protein seeding and disease propagation has received increasing attention due to the discovery of Lewy body formation in embryonic grafts of patients with Parkinson`s disease (PD) [6]. Similar mechanisms may also play a role in MSA although there are no convincing experimental or postmortem data [7]. Of note, cerebrospinal fluid (CSF) from patients with MSA seems capable of promoting in vitro α-synuclein fibril formation [8].

Genetics

Several independent research groups identified an association of single nucleotide polymorphisms (SNP) within the α-synuclein gene locus (SNCA) and increased MSA disease risk [9*, 10, 11]. However, these findings were not replicated in a Korean study due to high SNP frequencies in the control population [12]. Although a number of MSA pedigrees have been reported consistent with mendelian disease no genes have been identified to date [13, 14].

Clinical studies

The clinical presentation of MSA has been characterized in greater detail by a prospective EMSA study analyzing 437 MSA patients in 19 different European study sites [15**]. In accordance with previous studies, patients with MSA exhibited early and severe autonomic involvement with urinary disturbances, postural instability due to orthostatic hypotension and gastrointestinal symptoms [15**, 16]. Surprisingly, only one-third of patients with either documented orthostatic dysfunction or bladder disturbances received appropriate treatment [15**]. This finding emphasizes the urgent need for additional guidelines in this area. Dementia is regarded an exclusion criterion according to the consensus diagnostic criteria for MSA [17]. In contrast, data from a large prospective study assessing the natural history of MSA suggest that significant cognitive impairment develops in a substantial proportion of patients with MSA, often early in the disease [18]. Sleep disturbances may occur in up to 70% of MSA patients with rapid eye movement (REM) sleep behavior disorder (RBD) being the most common presentation [16]. Insomnia and excessive daytime sleepiness (EDS) occur less frequently (19% and 17%, respectively) [15**]. Additionally, up to 30% of patients have symptoms consistent with a restless legs syndrome [15**, 19*]. Interestingly, RBD may not only be a concomitant feature in the disease course but may also precede other clinical aspects by several years [20]. An intriguing polysomnographic study in levodopa-refractory MSA patients demonstrated improved facial expression, improved speech and limb movements during RBD [21], suggesting compensatory involvement of cerebellar and other non-basal ganglia motor circuits [21].

Depression, anxiety and pain are disabling symptoms of MSA. In a recent series 43% of MSA patients had probable depression according to the “Hospital Anxiety and Depression Scale” (HADS) [22]. In addition, this series revealed that more than 50% of patients with MSA suffer from anxiety [22].

Diagnostic work-up

The relevance of additional investigations in early suspected MSA remains unclear, however, neuroimaging appears to contribute most to diagnostic accuracy.

The predictive value of baseline putaminal abnormalities on 3-T MRI in differentiating early MSA-P from PD was assessed in a 3-year follow-up study [23]. A sum score based on putaminal abnormalities was highly specific (93%) at the expense of sensitivity (70.6%) [23]. In a voxel-based morphometry MRI approach, annual tissue-loss profiles revealed white matter reduction within the middle cerebellar peduncles in MSA patients consistent with degeneration of the ponto-cerebellar tract [24]. Additionally, white matter reduction along the corpus callosum at follow-up was reported suggesting a possible disease-specific pattern of neurodegeneration and cortical atrophy [24]. Diffusion tensor imaging (DTI) showed microstructural white matter abnormalities in patients with MSA-C suggesting that DTI may serve as a useful surrogate marker [25].

In a large cohort of patients with parkinsonian disorders a recent study showed high positive predictive values of automated metabolic expression patterns using fluorine-18-labelled-fluorodeoxyglucose positron emission tomography (FDG-PET). MSA patients were classified with a sensitivity of 85% and a specificity of 96% reflecting a 97% positive predictive value (PPV) and an 83% negative predictive value (NPV) [26*]. Similarly, a computer-assisted approach following an automated observer-independent algorithm was superior to a rater-based approach in classifying a sup¹²³I-β-CIT-SPECT image to either PD, MSA, PSP and healthy age-matched controls [27]. These findings suggest that automated image-based classification may help in discriminating MSA, PSP and PD in early-stage patients. As mentioned previously RBD may precede neurodegenerative disorders associated with loss of nigral dopaminergic neurons. In these patients, ¹²³I-2β-carbomethoxy-3β-(4-iodophenyl)-N-(3-fluoropropyl)-nortropane (¹²³I-FP-CIT) SPECT and transcranial sonography (TCS) were able to detect subclinical changes (i.e. decreased striatal ¹²³I-FP-CIT binding and substantia nigra hyperechogenicity) which might be useful markers to identify individuals at increased risk for development of PD, DLB or MSA [28]. A PET study assessing motor activation in MSA compared to PD showed activated cerebellar pathways in PD patients whereas MSA patients recruited frontoparietal cortical areas, likely reflecting compensation of basal-ganglia and cerebellar dysfunction [29]. ¹²³I-Meta-iodobenzylguanidine (MIBG) myocardial scintigraphy has been proposed as a useful tool to separate MSA from PD [30]. However, a recent longitudinal study showed decreased cardiac MIBG uptake in up to 30% of MSA patients irrespective of disease duration or severity [31]. Concurrent sympathetic Lewy body pathology may rarely accelerate cardiac denervation in such MSA patients [32]. Widespread brainstem, white matter and cortical α-synuclein deposition correlating with GCI density at postmortem was reported in 8 MSA patients using PET and ¹¹C-2-[2-(2-dimethylaminothiazol-5-yl)ethenyl]-6-[2-(fluoro)ethoxy]benzoxazole [33*], suggesting that in-vivo synuclein PET imaging may represent a novel surrogate marker of MSA.

The work-up of autonomic failure in MSA is mostly based on cardiac autonomic, neurourological and sudomotor tests. In a prospective study assessing the discriminative value of multiple cardiac autonomic and sudomotor indices demonstrated that MSA and PD may be separated by the severity, distribution, and pattern of autonomic deficits – even at study entry [34]. Multiple sensor pressure transducer measurements of bladder and urethral pressure characterized detrusorurethral dyssynergia in MSA patients more accurately than a single global measurement in a recent prospective study. Detrusor-underactivity was noted in 58% of MSA patients within 4 years of disease-onset and in 76% of patients thereafter [35].

Treatment

Symptomatic therapy is often ineffective in MSA. More recently, droxidopa, a precursor of noradrenaline also known as L-threo-dops, has been investigated in phase 3 trials for neurogenic orthostatic hypotension (NOH) including MSA (NCT00782340). The results are not available yet, however, phase 2 results appear promising [36, 37]. Unfortunately, disease-modifying interventions are not available for MSA, although there are increasing efforts to identify candidate neuroprotective agents in the preclinical testbeds that have become available [2].

A recent clinical trial assessing the efficacy of minocycline in patients with MSA failed to demonstrate beneficial effects [38*]. Despite this disappointing finding, research into disease-modifying strategies has accelerated. Park et al. [39] reported evidence of neuroprotection mediated by mesenchymal stem cells (MSC) in the double toxin-double lesion MSA-P model. A prior open label MSC trial in a series of MSA-C patients claimed a disease-modifying effect [40]. Erythropoietin improved behavioural deficits and attenuated neurodegeneration in transgenic MSA mice with GCI-like inclusions and neurodegeneration [41]. Intriguingly, reduced expression of oligodendroglial GDNF was recently reported in cerebellar and frontal white matter of MSA brains as well as in transgenic MSA mice highlighting a further potential therapeutic target [4]. Currently, a number of phase 2 safety and efficacy trials are ongoing including rasagiline (NCT00977665), rifampicin (NCT01287221), lithium (NCT00997672), intravenous immunoglobulins (NCT00750867), and autologous MSC (NCT00911365).

Dementia With Lewy Bodies

Severe cognitive decline frequently emerges many years after the onset of parkinsonism in patients with PD dementia (PDD) [42]. In contrast, dementia with Lewy bodies (DLB), i.e. dementia prior to the onset of parkinsonism [43], appears to be less common in movement disorder clinics. Both PDD and DLB share a similar distribution of Lewy bodies and neuronal loss.

Molecular pathogenesis

A recent study showed that more than 90% of α-synuclein aggregates in DLB cases were located in small deposits at the presynapses [44], suggesting that α-synuclein associated synaptic dysfunction drives neurodegeneration in DLB [45*]. Less is known about interactions between α-synuclein and other neuronal proteins, however, α-synuclein was shown to promote tau accumulation in vivo [46]. Intriguingly, a combined transgenic mouse model expressing human APP, human tau and human α-synuclein revealed that amyloid-β, tau, and α-synuclein interact synergistically in vivo, causing accelerated cognitive decline [47].

Genetics

DLB appears to be a sporadic disorder. Familial aggregation has rarely been reported, however, sequencing of candidate genes in a Belgian DLB pedigree failed to identify a pathogenic mutation [48]. A recent case-control study demonstrated that siblings of clinically diagnosed DLB patients are at increased risk of developing the disease [49].

Clinicopathological studies

There is considerable overlap between PDD and DLB [43], however, a study of 30 brains with Lewy body dementia revealed that only 2 of 13 PDD cases (15%) showed AD pathology in midbrain areas, whereas 12 of 17 DLB cases (71%) exhibited midbrain involvement [50]. Moreover, 4 of the DLB cases (24%) but none of the PDD cases exhibited amyloid pathology in the cerebellum [50]. Additionally, in a recent clinicopathological study all patients with mild cognitive impairment and RBD exhibited post-mortem confirmed Lewy body disease, with involvement of brainstem and cortical areas [51].

Diagnostic work-up

A recent DTI study in DLB patients showed increased amygdalar diffusivity in the absence of tissue loss, consistent with microvacuolation suggesting that DTI may be helpful in separating DLB from AD [52]. Cholinergic deficits seem to be important in the development of DLB as shown in a neuroimaging study analysing neurotransmitter changes using a voxel-based statistical parametric mapping and region of interest-based statistics [53*]. Moreover, this study demonstrated that both DLB and PDD patients share the same dopaminergic and cholinergic deficit profile in the brain [53*]. A meta-analysis determined confirmed the utility of cardiac MIBG scintigraphy as an accurate diagnostic marker separating DLB from other dementias with a pooled sensitivity of 98% and pooled specificity of 94% [54].

Treatment

A large prospective randomized placebo-controlled study reported significant improvement of clinical global impression of change in patients with either DLB or PDD with benefits being more prominent in the PDD subgroup [55*]. While memantine-mediated global improvement was replicated in mild to moderate DLB patients of another large multi-center trial, no beneficial effects were observed in the PDD group [56*].

Tauopathies

PSP and CBD are rare parkinsonian disorders resulting from pathological tau aggregation firmly placing these disorders within the enlarging spectrum of tauopathies including the frontotemporal dementias.

Progressive Supranuclear Palsy

During the last years infrequent variants due to PSP tau pathology have been separated from the “classical” syndrome (Richardson syndrome, RS) [57*]. These variants that challenge the concept of clinicopathologic PSP [58] include PSP-parkinsonism (PSP-P) [58], pure akinesia with gait freezing (PAGF) [59], CBS [60**, 61*], and progressive non-fluent aphasia (PSP-PNFA) [62].

Molecular pathogenesis

Data from human PSP brains and experimental models suggest that tau is hyperphosphorylated by kinases, leading to detachment from microtubules. Unbound phosphorylated tau, particularly as the 4R isoform, has an intrinsic propensity to form cytotoxic aggregates [63**]. Microtubules, in turn, are destabilized after detachment of tau and become depolymerized to tubulin-monomers, thereby losing their vital functions. Mitochondrial dysfunction appears to accelerate pathological tau aggregation [63**].

Genetics

Based on the current literature and with few exceptions, PSP can be considered a sporadic disorder. Vidal et al. [64] recently reported lack of familial aggregation 79 pairs of PSP cases and one first-degree relative matched for age, gender and living area for each PSP case. The pathogenic importance of tau aggregates in PSP is underpinned by the robust association of the H1 tau haplotype with PSP [65]. A recent genome-wide association study (GWAS) reported a major locus on chromosome 11p12-p11 which correlated with PSP disease risk [66]. Familial PSP is rare [67] and sometimes related to familial frontotemporal dementia with parkinsonism linked to chromosome 17 (FTDP-17) [68].

Clinical studies

A prospective natural history study determined a median survival of 8.0 years in a series of 197 PSP patients consistent with the rapid progression of this devastating disorder [69]. Male gender, older age at disease-onset and higher PSP rating scale scores were independent predictors of shorter survival [69]. A natural history study confirmed previous observations of high levels of cognitive impairment associated with PSP occurring in up to 50% of patients at early disease stages [18]. Furthermore, the cognitive profile characterized by frontal-executive dysfunction was similar to a control cohort of MSA patients [18]. Interestingly, impulse control disorders (ICDs) emerged during dopamine agonist treatment in 3 cases of pathologically proven PSP indicating that dopamine agonist induced behavioural side effects may not be limited to PD [70].

Clinicopathological studies

Due to the lack of specific biomarkers it has been suggested that the recently recognized clinical variants may be included under the generic term PSP reflecting rapidly progressive disease and similar pathology including 4R tau inclusions, tufted astrocytes, and coiled bodies [57*]. The regional differences in pathological severity almost certainly account for the clinical differences and logically correlate with the different clinical features. The NINDS-SPSP were established 15 years ago focusing on the classic RS presentation [71]. Several clinical features such as visual hallucinations, drug-induced dyskinesias and autonomic dysfunction are uncommon in PSP-P and discriminate this subtype from other parkinsonian disorders including PD, DLB and MSA [72].

Diagnostic work-up

The diagnosis of PSP is primarily clinical. There is no consensus concerning additional investigations, however, neuropsychological testing and neuroimaging are often helpful [73, 74].

The diagnostic accuracy of brainstem MRI measurements such as the ratio between midbrain and pons areas (m/p-ratio), that between superior (SCP) and middle cerebellar peduncle (MCP) widths, and the MR parkinsonism index ([pons/midbrain]*[MCP/SCP], MRPI) was assessed in a study comparing different variants of PSP (i.e. PSP-P and PSP-RS) and PD [75]. While all measures were able to appropriately differentiate PSP-RS from PD, the m/p-ratio was the only measure found to distinguish PSP-P from PD [75]. Similarly, in a large cohort of 123 patients with neurodegenerative parkinsonian disorders including PSP, PD, and MSA-P, PSP cases had significantly higher MRPI values and smaller m/p-ratios compared to other groups [76*]. Furthermore, a voxel-based morphometry MRI approach found major anatomical differences between PSP-RS and PSP-P with additional grey matter loss within midbrain regions of patients with PSP-RS [77]. This finding was further supported by a study revealing thalamocortical atrophy as a defining feature of PSPRS, however, there was no correlation between atrophy and the presence of any specific cardinal clinical feature [78]. Microstructural brain pathology in line with neuropathological studies was demonstrated in PSP patients by diffusion tensor imaging (DTI) subjected to tract-based spatial statistics analysis [79*]. A recent functional imaging study revealed impaired subcortical and brainstem cholinergic neurotransmission in PSP consistent with degeneration of cholinergic nuclei such as PPN [80]. The diagnostic relevance of these findings remains to be established in future studies. As previously mentioned, a computer-assisted image analysis was superior to a rater-based approach in classifying a β-CIT-SPECT image to either PD, MSA, PSP or healthy age-matched controls [27].

The role of CSF or blood markers is controversial. Aerts et al. [81] reported that CSF α-synuclein and tau levels did not differentiate PSP from PD and other APDs. However, a low ratio of a truncated and extended tau form in the CSF distinguished PSP from related disorders [82].

Treatment

Thus far, symptomatic therapy is limited and there are no established disease-modifying interventions. In this situation, palliative measures including physical, occupational and speech therapy remain the most important aspects in the management of patients with PSP. Recent insights into pathogenic mechanisms of PSP allowed the development and preclinical testing of promising compounds targeting key pathogenic events.

Experimental data in tauopathy models suggest that the inhibition of GSK-3β might be a potential therapeutic target in PSP [63**]. Presently, several clinical trials with GSK-3β inhibitors such as lithium, valproic acid or NP031112 have been initiated or are ongoing in patients with PSP. The clinical study with lithium was terminated because of adverse events. NAPVSIPQ (davunetide) is a neuroprotective peptide interacting with the microtubule cytoskeleton to improve microtubule function [83]. A clinical trial to test the effects of danuvetide in PSP is presently ongoing (NCT01056965). To improve mitochondrial dysfunction the short-term effects of coenzyme Q10 were studied in 21 PSP patients. Compared to placebo, this treatment significantly improved cerebral metabolism as measured by ³¹P- and ¹H-MRS [84]. Several clinical measures including the PSP rating scale and the frontal assessment battery also improved. Whether long-term administration of coenzyme Q may impact on disease progression in PSP is currently investigated in a phase 3 trial (NCT00382824).

Corticobasal Degeneration

CBD is the least common neurodegenerative APD and may only be diagnosed post-mortem. The clinical presentation of CBD is variable and includes the CBS, PSP, progressive aphasia and frontotemporal dementia. However, CBD research is hampered by the rarity of the disorder patients being scattered over a large geographic area [1**].

Molecular pathogenesis

Multiple pathogenic mechanisms have been proposed in tauopathies in general, including mitochondrial dysfunction, oxidative stress, abnormal axonal transport, inflammation, and particularly disturbances in tau metabolism [85]. However, due to the complexity and rarity of CBD there have not been major recent advances.

Genetics

A recent study showed that the inheritance of the tau H1/H1 haplotype is associated with the severity of motor function but not age at onset and cognition in CBS patients [86]. Moreover, there was a trend towards shortened survival in the H1/H1 group which failed to reach significance, most likely due to sample size limitations [86]. Finally, frontotemporal dementia patients with defined tau mutations may also present with a CBD phenotype [87].

Clinicopathological studies

CBD is characterized by deposits of 4-repeat tau in cortical and striatal neurons and glia forming astrocytic plaques and by thread-like lesions affecting white and gray matter [88]. A recent clinicopathological study showed that only the minority of post-mortem confirmed CBD cases were diagnosed correctly in life (5 out of 19 cases) and that the clinical syndrome of CBS is not specific for CBD [60**]. Intriguingly, most of the pathologically confirmed CBD cases were characterized clinically by a PSP-like presentation with delayed onset of vertical supranuclear gaze palsy [60**]. TAR-DNA-binding protein 43 (TDP-43) inclusions that are distinctive of frontotemporal lobar degeneration with ubiquitin-positive inclusions (FTLD-U) may occur in a minority of CBD and PSP cases [89, 90], highlighting the overlap between CBD, PSP and FTLD-U.

Diagnostic work-up

Elevated levels of tau and phosphorylated tau protein may be found in patients with CBS, however, diagnostic accuracy seems to be poor [91]. Intriguingly, an AD-like CSF pattern was associated with early memory decline in CBS patients according to a recent study [92]. Voxel-based morphometry may be used to predict underlying pathology in patients presenting with CBS as nicely shown by Whitwell and coworkers in a series of 24 patients [61*]. While widespread atrophy was associated with either FTLD or AD, CBD or PSP pathologies were characterized by focal atrophy predominantly involving premotor and supplemental motor areas [61*]. Cholinergic imaging using ¹¹C-N-methylpiperidin-4-yl acetate PET revealed decreased acetylcholinesterase activity in CBS and PSP patients, with CBS predominantly involving the paracentral region as well as frontal, parietal and occipital cortices [93].

Treatment

Since symptomatic therapy is largely ineffective in CBS patients there have been limited efforts to evaluate disease-modifying interventions. Lithium, a glycogen synthase kinase 3-β (GSK-3β) inhibitor, ameliorated disease progression in transgenic tauopathy models [94] and is currently evaluated in a Phase II study combining patients with CBS and PSP (NCT00703677). In addition, the effects of NAPVSIPQ (davunetide) [95], a neuroprotective peptide that stimulates microtubule assembly and reduces tau phosphorylation are currently investigated in CBS (NCT01056965). Until specific biomarkers for underlying CBS pathologies are available recruitment of such patients into clinical trials seems warranted considering that most such patients will have an underlying tauopathy.

Conclusion and Future Perspectives

In recent years APDs have been increasingly recognized as important causes of parkinsonism and, thus, research into underlying pathogenic mechanisms, the aetiopathogenesis, diagnostic and progression markers, and novel interventional therapies has intensified. During the next few years we will see a further acceleration of research into genetics, molecular biology and neuroimaging of these disorders which will allow the development and assessment of disease-modifying strategies targeting key pathogenic events.

Key Points.

• Important genetic advances in APDs during the last years include the discovery of variants in the synuclein and tau gene with increased disease risk for MSA and PSP.

• The diagnostic accuracy and surrogate monitoring of disease activity has improved with the aid of advanced neuroimaging tools.

• Preclinical evidence, particularly in MSA models, stimulates further research into disease-modifying strategies in APD patients.