Parkinson’s disease (PD) is characterized by bradykinesia, rigidity, sometimes rest tremor and postural instability. A disturbed α-synuclein protein forming so-called Lewy Bodies seems to play a causal role, which was a reason to designate PD as a α-synucleinopathy. The main pathophysiological changes result from degeneration of catecholaminergic, especially dopaminergic cells in brainstem regions.
A characteristic metabolic covariance pattern has been identified in PD patients (PD-related pattern, PDRP) showing regionally relatively increased metabolism in the globus pallidus and putamen, thalamus, pons and cerebellum and relatively decreased metabolism in the lateral frontal, premotor and parietal association areas (Ma, et al. 2007b). Network expression in PD patients also increases linearly with disease progression (Huang, et al. 2007b). Tang et al. tried to study network changes in the PD-related motor pattern before symptom onset by studying 15 hemiparkinsonian patients and focusing mainly on the “presymptomatic” hemisphere. They conclude that abnormal PDRP activity antecedes the appearance of motor signs by approximately 2 years (Tang, et al. 2010a). However, this needs to be proven in future research in true presymptomatic patients.
In addition to motor symptoms, cognitive dysfunction is also common in PD, especially executive and visuospatial dysfunction. FDG PET studies have been performed to study these specific symptoms and their relations with neural network pathophysiology. The Eidelberg research group has shown PD subclassifications related to specific symptoms. Network analysis with the SSM/PCA approach detected a significant covariance pattern in non-demented PD patients that correlated with memory and executive functioning tasks. The expression of this PD-related cognitive pattern (PDCP) in individual patients correlated with severity of cognitive dysfunction (Huang, et al. 2007a).
Multiple system atrophy (MSA) is a sporadic neurodegenerative brain disease which affects both men and women and generally starts in the sixth decade of life. The main clinical features are parkinsonism, autonomic failure, cerebellar ataxia, and pyramidal signs in any combination. However, two major motor presentations can be distinguished. Parkinsonian features predominate in 80% of patients (MSA-P subtype) and cerebellar ataxia is the main motor feature in 20% of patients (MSA-C subtype) (Gilman, et al. 2008, Wenning, et al. 1997).
In MSA-P the striatonigral system is the main site of pathology but less severe degeneration can be widespread and normally includes the olivopontocerebellar system. In MSA-C pathological changes are mainly seen in the olivopontocerebellar system and involvement of striatum and substantia nigra are less severe (Wenning, et al. 1997). The discovery of glial cytoplasmic inclusions in MSA brains highlighted the unique glial pathology as biological hallmark of the disease. Their distribution selectively involves basal ganglia, supplementary and primary motor cortex, the reticular formation and pontocerebellar system. Glial cytoplasmic inclusions contain besides classical cytoskeletal antigens also α-synuclein, which is a presynaptic protein present in Lewy Bodies, and this accumulation seems to play a central part not only in MSA but also in other α-synucleinopathies such as PD and DLB.
Disease-related metabolic patterns were also present in MSA consisting of hypometabolism in putamen and cerebellum in MSA (Eckert, et al. 2008). Poston et al. found that differences in expression of the MSA-related pattern correlated with clinical disability (Poston, et al. 2012).
The clinical picture of progressive supranuclear palsy (PSP) has been first described by Steele, Richardson and Olszewski (Steele JC, Richardson J, Olszewski J. 1964) and is characterized by progressive parkinsonism, early gait and balance impairment, vertical gaze palsy and more profound frontal cognitive disturbances. PSP is one of several neurodegenerative diseases characterised by accumulation of hyperphosphorylated tau (tauopathy), forming abnormal filamentous inclusions in neurons and glia in the precentral and postcentral cortical areas but also in the thalamus, subthalamic nucleus, red nucleus and substantia nigra. Other neurodegenerative brain diseases which show disturbances in tau protein handling are corticobasal degeneration (CBD) and frontotemporal dementia (FTD) but there is also overlap in pathology with Alzheimer’s disease (AD).
However the metabolic brain patterns in these tauopathies are quite different. The covariance pattern of PSP consists of decreased metabolism in the prefrontal cortex, frontal eye fields, caudate nuclei, medial thalamus and upper brainstem (Eckert, et al. 2008). Brain stem atrophy and atrophy of the medial frontal cortical regions have also been reported in histopathological studies (Hauw, et al. 1994).
The most striking features of patients with corticobasal degeneration (CBD) include marked asymmetrical parkinsonism and apraxia but also postural instability, limb dystonia, cortical sensory loss, dementia and the alien limb phenomenon. CBD is one of the tauopathies and
clinical diagnosis is complicated by both the variability of presentation of true CBD and the syndromes that look alike but are caused by other tauopathies with parkinsonism like PSP or FTD (Josephs, et al. 2006). However with functional neuroimaging a clear distinction can be made. In CBD a typical pattern of hypometabolism is seen in cortical regions contralateral to the affected body side, including parieto-temporal regions, prefrontal cortex and motor cortex. Furthermore, a decrease can be found in the contralateral caudate nucleus, putamen and thalamus (Eckert, et al. 2005, Teune, et al. 2010). No covariance pattern has been described using the SSM/PCA technique in CBD.
The clinical overlap of dementia and parkinsonism is highlighted in dementia with Lewy Bodies (DLB). These patients show besides dementia extrapyramidal motor symptoms and marked neuropsychiatric disturbances including visual hallucinations, depression, variability in arousal and attention (McKeith. 2006). Consistent observation of a metabolic reduction in the medial occipital cortex in DLB patients (Minoshima, et al. 2001, Teune, et al. 2010) using FDG PET imaging suggests the use of FDG PET in the differential diagnosis of AD and DLB and of PD and DLB. Minoshima et al. found that the presence of occipital hypometabolism distinguished DLB from AD with 90% sensitivity and 80% specificity when using post-mortem diagnosis as the gold standard diagnosis (Minoshima, et al. 2001).
Alzheimer’s disease (AD) is a progressive neurodegenerative brain disease accounting for 50-60% of cases of dementia. AD is characterized by a severe decline in episodic memory together with general cognitive symptoms such as impaired judgement, decision making and orientation (McKhann, et al. 1984). A correct clinical diagnosis can be difficult, especially in early disease stages or in patients with for example comorbid depression, high education or young age (Bohnen, et al. 2012). FDG PET imaging can be used to assist in the differential diagnosis, because for different dementia syndromes, a separate pattern of hypometabolism can be found. In Alzheimer’s disease (AD), decline of FDG uptake in posterior cingulate, temporoparietal and prefrontal association cortex was related to dementia severity (Herholz, et al. 2002). Foster et al used visual interpretation of an automated three-dimensional stereotactic surface projection technique of patients with AD and FTD. They showed that visual interpretation of FDG PET scans after training is more reliable and accurate in distinguishing FTD from AD than clinical methods alone (Foster, et al. 2007).
Although multivariate analytical techniques might identify diagnostic patterns that are not captured by univariate methods, they have rarely been used to study neural correlates of Alzheimer’s Disease or cognitive impairment. Because cognitive processes are the result of integrated activity in networks rather than activity of any one area in isolation, functional connectivity can be better captured by multivariate methods. A study from Habeck et al. examined the efficacy of multivariate and univariate analytical methods and concluded that multivariate analysis might be more sensitive than univariate analysis for the diagnosis of early Alzheimer’s disease (Habeck, et al. 2008). Our research group has very recently used a multivariate analytical technique (SSM/PCA) to identify a metabolic brain pattern for Alzheimer’s disease (Teune et al, 2013, under review). This AD-related covariance pattern was not surprisingly characterized by bilateral relatively decreased metabolic activity in the bilateral temporal regions, precuneus, posterior cingulated en angular gyrus, the inferior parietal region and supramarginalis. Relatively increased metabolic activity was seen in the subcortical white matter, cerebellum and sensorimotor cortex.
Frontotemporal dementia (FTD) is one of the neurodegenerative diseases commonly mistaken for AD. FTD patients do not have a true amnestic syndrome but can present with either gradual and progressive changes in behaviour, or gradual and progressive language dysfunction. Gross examination of the post-mortem brain from a patient with FTD usually reveals frontal or temporal lobar atrophy or both, but the distribution or severity of brain atrophy are not specific for a particular neurodegenerative brain disease. Jeong et al. and Diehl-Schmid et al. analysed FDG PET scans of FTD patients on a voxel-by-voxel basis using Statistical Parametric Mapping (SPM). They found hypometabolism depending on disease stage in the frontal lobe, parietal and temporal cortices (Diehl-Schmid, et al. 2007, Jeong, et al. 2005).