The application of network mapping in differential diagnosis of parkinsonian disorders

Clinical Neuroscience Research 6 (2007) 359–366 www.elsevier.com/locate/clires

The application of network mapping in differential diagnosis of parkinsonian disorders Thomas Eckert a

a,b,*

, Christine Edwards

a

Center for Neurosciences, Feinstein Institute for Medical Research, North Shore-Long Island Jewish Health System, 350 Community Drive, Manhasset, NY 11030, USA b Department of Neurology II, University of Magdeburg, Germany

Abstract Although approximately 1–3% of the population over age 65 have Parkinson’s disease (PD), only about 75% of the patients diagnosed with parkinsonism have PD. The differential diagnosis of parkinsonian disorders based on clinical symptoms alone is particularly difficult during the early stages of the disease. A number of imaging strategies have been developed to differentiate between these clinically similar conditions. The assessment of abnormal patterns of brain metabolism, either by visual inspection or using computer-assisted algorithms, can be used to discriminate between classical PD and atypical variant conditions such as multiple system atrophy (MSA), progressive supranuclear palsy (PSP), or corticobasal ganglionic degeneration (CBGD). Recent advances in network quantification routines have created the basis for fully automated differential diagnosis. Using PET, investigators have identified specific disease-related spatial covariance patterns that are characteristic of PD and its variants. By computing pattern expression in individual patient scans, it has become possible to determine the likelihood of a specific diagnosis. In this review, we describe the various imaging techniques that have been used to diagnose PD with emphasis on the application of network tools. Analogous methods may have value in the assessment of other neurodegenerative and neuropsychiatric conditions. Ó 2007 Published by Elsevier B.V. on behalf of Association for Research in Nervous and Mental Disease. Keywords: Parkinsonism; Positron emission tomography; Brain metabolism; Differential diagnosis; Network analysis; Spatial covariance patterns

1. Introduction Idiopathic Parkinson’s disease (PD) is a relatively common disorder with a prevalence of 1–3% in the population over 65 years of age [1]. Clinically, parkinsonism is characterized by the existence of at least two of the following motor symptoms: tremor, hypokinesia, and rigidity. Classical PD is differentiated from other parkinsonian disorders such as multiple system atrophy (MSA), progressive supranuclear palsy (PSP), corticobasal ganglionic degeneration (CBGD), and Lewy body dementia (DLB) by a combina*

Corresponding author. Address: Center for Neurosciences, Feinstein Institute for Medical Research, North Shore-Long Island Jewish Health System, 350 Community Drive, Manhasset, NY 11030, USA. Tel.: +1 516 562 1057; fax: +1 516 562 1008. E-mail address: [email protected] (T. Eckert).

tion of an asymmetry of symptoms, possible presence of a resting tremor, and a long-term responsiveness to levodopa. The differential diagnosis of parkinsonian disorders based solely on clinical symptoms remains unsatisfactory in that only 76% of patients thought to have PD prove to have this diagnosis at postmortem [2]. The overall assessment of a trained movement disorder specialist after clinical follow-up has been shown to provide the most accurate diagnosis [3]. However, differentiation of these disorders by clinical assessment alone can be difficult in the earliest stages of disease. Indeed, the clinical diagnosis made by a movement disorder specialist at the initial visit had to be changed in approximately one-third of patients by the fifth year of symptoms [3]. Early differential diagnosis in patients with parkinsonism is important for prognosis and the course of treatment. While the life expectancy of PD patients resembles that of

1566-2772/$ - see front matter Ó 2007 Published by Elsevier B.V. on behalf of Association for Research in Nervous and Mental Disease. doi:10.1016/j.cnr.2007.05.001

360

T. Eckert, C. Edwards / Clinical Neuroscience Research 6 (2007) 359–366

their healthy contemporaries, that of patients with atypical parkinsonian syndromes such as MSA, PSP, and CBGD, is markedly reduced. Moreover, patients with atypical parkinsonian syndromes generally do not benefit from pharmacologic and surgical interventions, although they can potentially experience the side effects of these therapies [4]. Accurate differential diagnosis of parkinsonian syndromes at the earliest stages of disease is also important for the optimal assessment of disease modifying agents. Newly diagnosed, de novo PD patients are often enrolled in these treatment trials. However, specific signs differentiating PD and parkinsonian syndromes (e.g., responsiveness to dopaminergic therapy) are often not well articulated at early clinical stages of disease. The enrollment of such patients might result in inaccurate study results. Considering the importance of early differential diagnosis, it is not surprising that a number of neuroimaging techniques have been developed to improve the accuracy of discriminating between PD and its clinical look-alikes. Widely available neuroimaging modalities such as position emission tomography (PET) and single photon emission computed tomography (SPECT) have provided novel insights into the pathophysiology of PD and other parkinsonian syndromes [5]. In PD, radiotracer-based imaging studies have been applied to assess nigrostriatal presynaptic function. Functional imaging techniques have also been used to study neuronal activity by quantifying regional glucose metabolism and cerebral blood flow in the resting condition. These assessments have contributed considerably to the understanding of abnormal neuronal circuitry underlying the pathophysiology of parkinsonism [6]. Other imaging methods used for differential diagnosis include cardiac sympathetic denervation and various magnetic resonance imaging (MRI) techniques. In this review, we will discuss the relative advantages and shortcomings of radiotracer-based imaging techniques in the clinical diagnosis and management of PD as well as the potential use of network mapping techniques to differentiate between PD and its atypical variants. 2. Imaging dopaminergic function in parkinsonism Neurodegenerative diseases, particularly those affecting the basal ganglia and related pathways, are often associated with the loss of nigrostriatal dopaminergic projections. The integrity of this system can be assessed by neuroimaging methods utilizing radioligands that bind to pre- or postsynaptic components. On the other hand, the functional status of the basal ganglia efferent projections can be assessed using imaging to assess regional cerebral blood flow and metabolism as measures of neural activity. 2.1. Presynaptic dopaminergic imaging The most commonly applied PET radiotracer to assess the integrity of presynaptic nigrostriatal dopaminergic nerve terminals is [18F]fluorodopa (FDOPA) (see [7]).

PET studies with this tracer measure the rate of decarboxylation of [18F]fluorodopa to [18F]fluorodopamine by dopa decarboxylase (DDC) and its subsequent storage in the striatal dopaminergic nerve terminals. FDOPA PET scans have routinely been analyzed by a multiple time graphical approach incorporating plasma or brain input functions [8], or by formal compartmental models to estimate the specific rate constant for striatal DDC [9,10]. Subsequent studies revealed that the striatooccipital ratio (SOR) for FDOPA determined from a single 10-min 3D PET scan can be as accurate as the more complex measures [8,11,12]. FDOPA PET measurements have been used to obtain objective correlates of disease severity and to discriminate early stage PD patients from patients with essential tremor, psychogenic movement disorders, and healthy control subjects (for review see [5]). FDOPA PET has also been used to differentiate among parkinsonian syndromes [13]. In patients with early stage PD, FDOPA uptake is relatively preserved in the caudate and anterior putamen. By contrast, in patients with atypical parkinsonian syndromes such as MSA, equivalent impairment of FDOPA uptake can be observed in the caudate and the putamen [14]. However, these dopaminergic signatures are often insufficient to discriminate PD from atypical parkinsonism at early clinical stages [15]. Indeed, striatal FDOPA uptake can be reduced in other parkinsonian movement disorders such as MSA, PSP, Wilson’s disease [16], Guamanian amyotrophic lateral sclerosis (ALS– PD complex; [17]), and X-linked Filippino dystonia parkinsonism [18]. Additionally, asymmetrical parkinsonian syndromes such as HPHA and CBGD can show relative reductions in basal ganglia FDOPA uptake contralateral to the affected side [19,20]. 2.1.1. DAT binding The development of radiotracers that bind to the striatal dopamine transporter (DAT) has led to another means for directly imaging the nigrostriatal dopaminergic system with PET or SPECT. DAT enables the release and reabsorption of dopamine in the nigrostriatal intersynaptic cleft. A number of radiotracers, mostly cocaine analogs, have been developed to quantify striatal DAT binding as an objective marker of the integrity of presynaptic nigrostriatal dopamine terminals (see [7], for review). The most extensively studied agents in this category are the cocaine analogues, such as 2-b-carbomethyl-3b-(4-iodophenyl) tropane (bCIT) and its fluoroalkyl esters [21]. Striatal DAT binding can be quantified by any of these agents, which differ primarily in the time of attained equilibrium and the duration of the scanning procedure. The major application of DAT binding imaging procedures has been to assess the rate of decline in presynaptic dopaminergic function in PD patients. In SPECT studies using [123I]bCIT [22], striatal DAT binding declined by 7.1% per year in short duration PD, and 14.9% in short duration atypical parkinsonian syndromes. The rate of decline in PD appears to slow markedly with evolving disease [22]. In

T. Eckert, C. Edwards / Clinical Neuroscience Research 6 (2007) 359–366

another bCIT study, 32 PD patients and 24 healthy controls underwent longitudinal SPECT imaging over a 4-year period. PD subjects demonstrated a decline in striatal DAT binding of approximately 11.2% per year from baseline as compared to 0.8% per year in control subjects [23]. A number of DAT binding studies have been performed to assess the use of this method in differentiating parkinsonian syndromes. Booij and colleagues used [123I]FPCIT SPECT to differentiate idiopathic PD patients from those with other forms of parkinsonism [24]. In a study of 157 PD and 26 MSA patients who underwent bCIT SPECT, striatal uptake was markedly reduced in both the PD and MSA groups. However, the MSA subjects displayed more symmetric DAT loss compared with the PD patients, in accord with their clinical manifestations. Comparison of the relative loss of DAT binding in the caudate and putamen did not improve diagnostic accuracy in distinguishing between PD and MSA [25]. A more recent study utilizing bCIT SPECT revealed marked reductions in the midbrain of MSA patients [26]. These reductions delineated MSA patients from PD patients and normal control subjects. However, these results could not be reproduced using the DAT binding ligand FPCIT [27]. It has been shown that striatal DAT binding decreases with age in healthy volunteers and PD patients [28–30]. Therefore, age correction is generally needed when comparing subjects of different age groups or when conducting longitudinal progression studies. 2.2. Postsynaptic dopaminergic imaging Dopamine receptor bearing neurons constitute approximately 80% of the neuronal population in the striatum [31,32]. Imaging studies using ligands that bind selectively to striatal D2 receptors offer a quantitative means to differentiate among parkinsonian syndromes. A SPECT study was performed using [123I]-(S)-5-iodo-7-N-[(1-ethyl-2-pyrrolidinyl)methyl] carboxamido-2,3-dihydrobenzofuran (IBF) to assess D2 receptor binding in patients with PD, MSA, and PSP [33]. D2 binding in posterior putamen was abnormally elevated in the untreated PD cohort, and significantly reduced in the MSA group. The ratio of posterior putamen to caudate D2 binding revealed the caudate to be greater in 16 of 18 PD patients and in all PSP patients; the caudate was lower in 5 of 7 MSA patients. These findings suggest that SPECT quantification of striatal D2 receptor binding can be helpful in discriminating MSA from other forms of parkinsonism. Similarly, [123I]iodolisuride was used with SPECT to discriminate PD from patients with atypical parkinsonian syndromes [34], reporting that the striato-occipital region was statistically different between these two patient groups. D2 receptor function in early stage PD has been assessed utilizing two PET ligands: [11C]raclopride (RAC) and [11C]-N-methylspiperone (NMSP) [35]. With both methods, receptor binding was increased in the putamen contralateral to the predominant symptoms as compared with the

361

ipsilateral side. Although D2-receptor binding may be upregulated in the early disease stages of PD, this compensatory function is lost with advancing disease, with a normalization or reduction of D2-binding during the later stages of PD [36]. 2.2.1. Metabolic imaging in parkinsonian disorders Besides quantifying dopaminergic function in neurodegenerative disorders, functional brain imaging can also be used to measure regional cerebral perfusion and glucose utilization as indices of local synaptic activity. The functional imaging approaches provide unique information regarding the activity of cortico-striato-pallido-thalamocortical (CSPTC) loops [37] in PD and related movement disorders (see [6,38], for reviews). 2.2.2. Use of computer-assisted methods for imaging-based diagnosis Metabolic imaging can provide a valuable tool to differentiate among parkinsonian syndromes. In a recent study, we defined metabolic imaging criteria for the diagnosis of PD, MSA, PSP, and CBGD with [18F]fluorodeoxyglucose (FDG) PET. Using statistical parametric mapping (SPM), we compared FDG PET images from groups of patients with different forms of parkinsonism and from healthy volunteer subjects (see Fig. 1). These analyses were the basis for the development of templates for computerassisted differential diagnosis [39]. The imaging characteristics that defined each disease are as follows: (1) metabolic decreases of the putamen and the cerebellum in MSA, (2) metabolic decreases of the midbrain and the midline frontal cortex in PSP, (3) asymmetrical metabolism of the cortex and the basal ganglia with a relative hypometabolism contralateral to clinically most affected side in CBGD, and (4) increased metabolism of the putamen, in the absence of defining characteristics of the other syndromes in PD. Single subject FDG PET scans were then compared to scans from 10 healthy control subjects using SPM99. Significant metabolic increases and decreases (p < 0.05) for each individual patient/subject was overlaid onto a MRI template image for further evaluation. A blinded reader then assessed these single subject statistical maps, which displayed significant metabolic deviations from normal. Employing the defining diagnostic imaging criteria and visual comparison to the characteristic disease templates, each single subject was categorized to be PD, MSA, PSP, CBGD, or a healthy normal subject. Imaging diagnosis was compared to the clinical diagnosis after an average follow-up time of about 2 years, which is considered to be the ‘‘gold standard’’ for the diagnosis of parkinsonian patients [3]. More than 90% of 157 subjects were correctly diagnosed. This approach might be regarded as a very basic method of pattern/network recognition. In a first step, the imaging software displays characteristic metabolic patterns/networks for different diseases. It is however the role of the assessor/reader to determine how a single case statis-

362

T. Eckert, C. Edwards / Clinical Neuroscience Research 6 (2007) 359–366

Fig. 1. Characteristic patterns of abnormal regional glucose metabolism in Parkinson’s disease (PD), multiple system atrophy (MSA), progressive supranuclear palsy (PSP), and corticobasal ganglionic degeneration (CBGD). The images were obtained using SPM to compare eight patients in each diagnostic category with ten healthy control subjects. [Hot colors (red) display relative increases and winter colors (blue) display relative decreases in glucose metabolism for the patient groups relative to controls.]

tical map corresponds with one of a group of available templates. Even though this approach proved to be very accurate for differential diagnosis, it is investigator-dependent. Moreover, this approach only assesses whether a certain pattern is expressed or not. It does not quantify the degree to which a certain characteristic is expressed. 2.2.3. Metabolic networks in parkinsonian disorders In order to identify specific metabolic brain networks associated with PD and other parkinsonian disorders, we have used FDG PET imaging in conjunction with a multivariate analysis of the regional data [40]. This approach is based upon principal component analysis (PCA), which

allows for the identification of disease-related spatial covariance patterns (i.e., brain networks) and the quantification of pattern expression (i.e., network activity) in individual subjects (cf. [41]). Utilizing this mathematical approach, we identified a specific regional metabolic network in PD patients scanned in the resting state [42,43]. The PD-related covariance pattern (PDRP) is characterized by pallidal and thalamic hypermetabolism associated with metabolic decrements in the lateral premotor cortex (PMC), the supplementary motor area (SMA), the dorsolateral prefrontal cortex (DLPFC) and the parieto-occipital association regions. This abnormal topography has since been validated in dif-

T. Eckert, C. Edwards / Clinical Neuroscience Research 6 (2007) 359–366

ferent ways: (1) PDRP activity reproducibility discriminates PD patients from controls in subsequent populations scanned with FDG PET [44–46], as well as ECD SPECT perfusion methods [47]; (2) it correlates with disease severity and duration [48–50]. Because the PDRP is already expressed during early stages of the disease [15], the assessment of the expression of this disease specific network might also be valuable for differential diagnosis of parkinsonian syndromes. Indeed, we have previously shown that PDRP expression in PD patients is significantly different from healthy control subjects [46,51] and patients with atypical parkinsonian disorders [15]. While earlier studies utilized a region-of-interest (ROI)based approach to quantify PDRP expression that was partly investigator-dependent, we have recently developed a voxel-based approach to network quantification [48,52]. In a recent study using this approach, we found that parkinsonian patients without a dopaminergic deficit on FDOPA PET do not have discernable PDRP expression on metabolic imaging [53]. Clinical follow-up conducted 2 years after imaging confirmed that none of these patients developed signs and symptoms of classical PD. Because parkinsonian syndromes have similar clinical features, some imaging characteristics of these syndromes might resemble each other. In fact, both PD and PSP dis-

363

play metabolic increases in the thalamus and decreased metabolism in frontal cortical regions [39]. In order to differentiate between these syndromes, it may be helpful to characterize disease-related networks for each of the diagnostic possibilities. Utilizing the same voxel-based network approach that we used to identify and validate the PDRP [52], we recently characterized spatial covariance patterns associated with MSA and PSP [41]. We employed a prospective single case method to quantify the expression of these disease-related patterns in individual scans of PD, MSA, and PSP patients as well as healthy control subjects. While all three patterns appeared to have a high sensitivity to identify the respective diseases, the specificity was relatively low resulting in a high rate of false positives (i.e., more than one pattern showed a high pattern expression score) [41]. This suggests that the quantification of characteristic networks identified by comparing images from patients and controls might not be sufficient for differential diagnostic purposes. As mentioned above, parkinsonian syndromes express similar clinical features, and can also share neuropathological features. To differentiate between such diseases, the identification of specific networks that highlight the differences between two pathological conditions might be necessary. For instance, such a ‘‘differentiating’’ pattern can be identified by applying spatial

Fig. 2. Disease-related spatial covariance patterns for PD, MSA, and PSP (see text). These abnormal metabolic patterns were identified using principal components analysis as described elsewhere [41]. [The color scale indicates network-related regional metabolic increases and decreases for the respective patterns.]

364

T. Eckert, C. Edwards / Clinical Neuroscience Research 6 (2007) 359–366

covariance analysis to FDG PET data from a combined group of parkinsonian patients with classical PD as well as an alternative diagnosis like PSP or MSA. As part of an ongoing study, we recently developed new disease-related patterns for MSA and PSP using FDG PET images from ten patients and ten control subjects (see Fig. 2). Additionally, we generated differential diagnostic patterns comparing PD vs. PSP, PD vs. MSA, and MSA vs. PSP. We then quantified the expression of these five patterns, in addition to the PDRP, in scans from 79 PD patients, 25 MSA patients, 20 PSP patients, and 22 healthy control subjects using a fully automated algorithm that was blind to clinical diagnosis. We found that 85% of these subjects were correctly classified based upon the respective network scores, without any form of reader input. Examples of this approach are provided in Fig. 3. This new method quantified the individual pattern scores for the PDRP, MSA-RP, and PSP-RP in each subject. All patterns revealed a high sensitivity to their respective diagnosis. However, the MSA-RP pattern was not as specific as the PDRP and PSP-RP. Indeed, increased expression of the MSA-RP was also present in patients with PSP. Nonetheless, we found that individual patients with elevations in both the MSA-RP and PSP-RP scores were accurately categorized as having PSP. By contrast, a diagnosis of MSA was more likely if the MSA-RP score was elevated in the presence of low PSP-RP and PDRP scores. Employing a completely automated algorithm to classify the scans according to pattern score can provide a fully investigator-independent method for differentiating

between parkinsonian disorders. Such an approach does not require extensive experience in reading PET images, which may have limited the utility of FDG PET imaging as a tool for the differential diagnosis of parkinsonian disorders. Disease-related networks can also be assessed using other forms of metabolic imaging. While SPECT and arterial spin labeled (ASL) MRI cannot be used to assess regional cerebral metabolism, both techniques can be used to quantify cerebral perfusion which can provide similar information. The PET-derived PDRP appears to be a metabolic marker that is also expressed in brain perfusion data. Ma et al. [52] recently quantified PDRP expression in FDG and 15O-water PET scans obtained in the same individuals. PDRP expression was highly intercorrelated (p < 0.001) across both imaging modalities [52]. Even though SPECT has a lower spatial resolution than PET, the PET-derived PDRP is also expressed in ECD SPECT scans of PD patients [47]. In a recent study, we used a fully automated voxel-based approach to quantify PDRP expression in the ECD SPECT scans of 35 PD patients, 15 disease severity and age-matched MSA patients and 35 age-matched healthy control subjects [54]. PDRP expression in the PD group was significantly increased compared to the MSA group (p < 0.001) and to the group of healthy control subjects (p < 0.001). Receiver–operator characteristic (ROC) analysis revealed that these network measures discriminated PD from MSA and healthy control subjects, with a high sensitivity and specificity. Similar discrimination of PD patients from controls has been reported

Fig. 3. Examples of the use of metabolic imaging to differentiate patients with parkinsonian syndromes. The first column gives the ultimate clinical diagnosis for these patients. The second column displays the results of single case SPM analysis in which significant changes (p < 0.05) were overlaid onto an MRI template (see e.g., [39]). These changes are described qualitatively in the third column. The fourth column reports the subject scores for the various disease-related patterns (PDRP, MSA-RP, and PSP-RP, see text) for these individual patients. For each case, the highest value for network expression, corresponding to the most likely diagnosis, is highlighted in yellow.

T. Eckert, C. Edwards / Clinical Neuroscience Research 6 (2007) 359–366

by quantifying PDRP activity in ASL MRI scans (D. Eidelberg, personal communication). 3. Future perspectives The assessment of disease-related metabolic networks for differential diagnostic purposes may also be applicable to other diseases. FDG PET is a sensitive method to display early neuropathologic changes in a number of neurological and neuropsychiatric diseases. We have identified disease-related spatial covariance patterns for other movement disorders such as Huntington’s disease [55], Tourette syndrome [56], and dopa-responsive dystonia [51]. Moreover, neurodegenerative diseases such as Alzheimer’s disease and fronto-temporal dementia have also shown to express characteristic metabolic changes [57] and network abnormalities [58]. Validation of disease-related networks in these syndromes may create the basis for investigatorindependent differential diagnosis, as has been developed in parkinsonism. The application of spatial covariance analysis in the differential diagnosis of parkinsonian disorders has thus far been limited to metabolic imaging techniques. However, characteristic patterns of neuronal loss are used to diagnose various neurodegenerative disorders upon autopsy [2,59,60]. A variety of imaging techniques have shown significant changes in parkinsonian disorders that correspond to the respective histopathologic features [61,62]. In the future, network analysis might be applied to techniques like magnetization transfer imaging that display such changes. Such approaches may further improve the accuracy of currently available diagnostic methods. 4. Summary The development of the application of characteristic disease networks to differentiate between PD and related disorders is a rather new area of investigation. Preliminary studies indicate the potential of this novel approach. Compared to classic imaging assessment that requires the extensive experience of clinician-readers, the assessment of characteristic disease networks is fully automated and may not suffer from the pitfalls of investigator-dependent methods. References [1] Moghal S, Rajput AH, D’Arcy C, Rajput R. Prevalence of movement disorders in elderly community residents. Neuroepidemiology 1994;13(4):175–8. [2] Hughes AJ, Daniel SE, Kilford L, Lees AJ. Accuracy of clinical diagnosis of idiopathic Parkinson’s disease: a clinico-pathological study of 100 cases. J Neurol Neurosurg Psychiatry 1992;55(3):181–4. [3] Hughes AJ, Daniel SE, Ben-Shlomo Y, Lees AJ. The accuracy of diagnosis of parkinsonian syndromes in a specialist movement disorder service. Brain 2002;125(Pt 4):861–70. [4] Tarsy D, Apetauerova D, Ryan P, Norregaard T. Adverse effects of subthalamic nucleus DBS in a patient with multiple system atrophy. Neurology 2003;61(2):247–9.

365

[5] Eckert T, Eidelberg D. The role of functional neuroimaging in the differential diagnosis of idiopathic Parkinson’s disease and multiple system atrophy. Clin Auton Res 2004;14(2):84–91. [6] Trosˇt M, Dhawan V, Feigin A, Eidelberg D. PET and SPECT. In: Beal MF, Lang A, Ludolph A, editors. Neurodegenerative diseases: neurobiology pathogenesis and therapeutics. Cambridge: Cambridge University Press; 2005. p. 290–300. [7] Dhawan V, Eidelberg D. PET imaging in Parkinson’s disease and other neurodegenerative disorders. In: Gilman S, editor. The neurobiology of disease. San Diego: Elsevier Academic Press; 2006. p. 821–8. [8] Ishikawa T, Dhawan V, Chaly T, Margouleff C, Robeson W, Dahl JR, et al. Clinical significance of striatal DOPA decarboxylase activity in Parkinson’s disease. J Nucl Med 1996;37(2):216–22. [9] Gjedde A, Reith J, Dyve S, Leger G, Guttman M, Diksic M, et al. Dopa decarboxylase activity of the living human brain. Proc Natl Acad Sci USA 1991;88(7):2721–5. [10] Dhawan V, Ishikawa T, Patlak C, Chaly T, Robeson W, Belakhlef A, et al. Combined FDOPA and 3OMFD PET studies in Parkinson’s disease. J Nucl Med 1996;37(2):209–16. [11] Dhawan V, Ma Y, Pillai V, Spetsieris P, Chaly T, Belakhlef A, et al. Comparative analysis of striatal FDOPA uptake in Parkinson’s disease: ratio method versus graphical approach. J Nucl Med 2002;43(10):1324–30. [12] Nakamura T, Dhawan V, Chaly T, Fukuda M, Ma Y, Breeze R, et al. Blinded positron emission tomography study of dopamine cell implantation for Parkinson’s disease. Ann Neurol 2001;50(2):181–7. [13] Eidelberg D, Edwards C, Mentis M, Dhawan V, Moeller J, editors. Movement disorders: Parkinson’s disease. San Diego: Academic Press; 2000. [14] Brooks DJ, Ibanez V, Sawle GV, Quinn N, Lees AJ, Mathias CJ, et al. Differing patterns of striatal 18F-dopa uptake in Parkinson’s disease, multiple system atrophy, and progressive supranuclear palsy. Ann Neurol 1990;28(4):547–55. [15] Eidelberg D, Moeller JR, Ishikawa T, Dhawan V, Spetsieris P, Chaly T, et al. Early differential diagnosis of Parkinson’s disease with 18Ffluorodeoxyglucose and positron emission tomography. Neurology 1995;45(11):1995–2004. [16] Snow BJ, Bhatt M, Martin WR, Li D, Calne DB. The nigrostriatal dopaminergic pathway in Wilson’s disease studied with positron emission tomography. J Neurol Neurosurg Psychiatry 1991;54(1):12–7. [17] Snow BJ, Peppard RF, Guttman M, Okada J, Martin WR, Steele J, et al. Positron emission tomographic scanning demonstrates a presynaptic dopaminergic lesion in Lytico-Bodig. The amyotrophic lateral sclerosis- parkinsonism–dementia complex of Guam. Arch Neurol 1990;47(8):870–84. [18] Eidelberg D, Takikawa S, Wilhelmsen K, Dhawan V, Chaly T, Robeson W, et al. Positron emission tomographic findings in Filipino X-linked dystonia-parkinsonism. Ann Neurol 1993;34(2):185–91. [19] Eidelberg D, Dhawan V, Moeller JR, Sidtis JJ, Ginos JZ, Strother SC, et al. The metabolic landscape of cortico-basal ganglionic degeneration: regional asymmetries studied with positron emission tomography. J Neurol Neurosurg Psychiatry 1991;54(10): 856–62. [20] Przedborski S, Giladi N, Takikawa S, Ishikawa T, Dhawan V, Spetsieris P, et al. Metabolic topography of the hemiparkinsonismhemiatrophy syndrome. Neurology 1994;44(9):1622–8. [21] Chaly T, Dhawan V, Kazumata K, Antonini A, Margouleff C, Dahl JR, et al. Radiosynthesis of [18F]-N-3-fluoropropyl-2-beta-carbomethoxy-3-beta-(4-iodophenyl) nortropane and the first human study with positron emission tomography. Nucl Med Biol 1996;23(8):999–1004. [22] Pirker W, Djamshidian S, Asenbaum S, Gerschlager W, Tribl G, Hoffmann M, et al. Progression of dopaminergic degeneration in Parkinson’s disease and atypical parkinsonism: a longitudinal betaCIT SPECT study. Mov Disord 2002;17(1):45–53.

366

T. Eckert, C. Edwards / Clinical Neuroscience Research 6 (2007) 359–366

[23] Marek K, Innis R, van Dyck C, Fussell B, Early M, Eberly S, et al. [123I]beta-CIT SPECT imaging assessment of the rate of Parkinson’s disease progression. Neurology 2001;57(11):2089–94. [24] Booij J, Speelman JD, Horstink MW, Wolters EC. The clinical benefit of imaging striatal dopamine transporters with [123I]FP-CIT SPET in differentiating patients with presynaptic parkinsonism from those with other forms of parkinsonism. Eur J Nucl Med 2001;28(3):266–72. [25] Varrone A, Marek KL, Jennings D, Innis RB, Seibyl JP. [(123)I]betaCIT SPECT imaging demonstrates reduced density of striatal dopamine transporters in Parkinson’s disease and multiple system atrophy. Mov Disord 2001;16(6):1023–32. [26] Scherfler C, Seppi K, Donnemiller E, Goebel G, Brenneis C, Virgolini I, et al. Voxel-wise analysis of [123I]beta-CIT SPECT differentiates the Parkinson variant of multiple system atrophy from idiopathic Parkinson’s disease. Brain 2005;128(Pt 7):1605–12. [27] Van Laere K, Casteels C, De Ceuninck L, Vanbilloen B, Maes A, Mortelmans L, et al. Dual-tracer dopamine transporter and perfusion SPECT in differential diagnosis of parkinsonism using template-based discriminant analysis. J Nucl Med 2006;47(3):384–92. [28] Dhawan V, Eidelberg D. SPECT imaging in Parkinson’s disease. Adv Neurol 2001;86:205–13. [29] Kazumata K, Dhawan V, Chaly T, Antonini A, Margouleff C, Belakhlef A, et al. Dopamine transporter imaging with fluorine-18FPCIT and PET. J Nucl Med 1998;39(9):1521–30. [30] Tissingh G, Bergmans P, Booij J, Winogrodzka A, Stoof JC, Wolters EC, et al. [123I]beta-CIT single-photon emission tomography in Parkinson’s disease reveals a smaller decline in dopamine transporters with age than in controls. Eur J Nucl Med 1997;24(9):1171–4. [31] Parent A, Hazrati LN. Functional anatomy of the basal ganglia. I. The cortico-basal ganglia-thalamo-cortical loop. Brain Res Brain Res Rev 1995;20(1):91–127. [32] Parent A, Hazrati L. Functional anatomy of the basal ganglia. II. The place of subthalamic nucleus and external pallidum in basal ganglia circuitry. Brain Res Brain Res Rev 1995;20(1):128–54. [33] Kim YJ, Ichise M, Ballinger JR, Vines D, Erami SS, Tatschida T, et al. Combination of dopamine transporter and D2 receptor SPECT in the diagnostic evaluation of PD, MSA, and PSP. Mov Disord 2002;17(2):303–12. [34] Prunier C, Tranquart F, Cottier JP, Giraudeau B, Chalon S, Guilloteau D, et al. Quantitative analysis of striatal dopamine D2 receptors with 123I-iodolisuride SPECT in degenerative extrapyramidal diseases. Nucl Med Commun 2001;22(11):1207–14. [35] Kaasinen V, Ruottinen HM, Nagren K, Lehikoinen P, Oikonen V, Rinne JO. Upregulation of putaminal dopamine D2 receptors in early Parkinson’s disease: a comparative PET study with [11C]raclopride and [11C]N-methylspiperone. J Nucl Med 2000;41(1):65–70. [36] Antonini A, Schwarz J, Oertel W, Pogarell O, Leenders K. Long-term changes of striatal dopamine D2 receptors in patients with Parkinson’s disease: a study with positron emission tomography and [11C]raclopride. Mov Disord 1997;12:33–8. [37] Wichmann T, DeLong MR. Functional and pathophysiological models of the basal ganglia. Curr Opin Neurobiol 1996;6(6):751–8. [38] Eckert T, Eidelberg D. Neuroimaging and therapeutics in movement disorders. Neurorx 2005;2(2):361–71. [39] Eckert T, Barnes A, Dhawan V, Frucht S, Gordon MF, Feigin AS, et al. FDG PET in the differential diagnosis of parkinsonian disorders. Neuroimage 2005;26(3):912–21. [40] Carbon M, Eidelberg D. Modulation of regional brain function by deep brain stimulation: studies with positron emission tomography. Curr Opin Neurol 2002;15(4):451–5. [41] Spetsieris P, Ma Y, Dhawan V, Moeller J, Eidelberg D. Highly automated computer-aided diagnosis of neurological disorders using functional brain imaging. In: Reinhardt JM, Pluim JPW, editors. Proceedings of SPIE – medical imaging. San Diego; 2006. p. M1–12. [42] Eidelberg D, Moeller JR, Dhawan V, Sidtis JJ, Ginos JZ, Strother SC, et al. The metabolic anatomy of Parkinson’s disease: comple-

[43]

[44]

[45]

[46]

[47]

[48]

[49]

[50]

[51]

[52]

[53]

[54]

[55]

[56]

[57]

[58]

[59]

[60]

[61]

[62]

mentary [18F]fluorodeoxyglucose and [18F]fluorodopa positron emission tomographic studies. Mov Disord 1990;5(3):203–13. Eidelberg D, Moeller JR, Dhawan V, Spetsieris P, Takikawa S, Ishikawa T, et al. The metabolic topography of parkinsonism. J Cereb Blood Flow Metab 1994;14(5):783–801. Feigin A, Ghilardi MF, Fukuda M, Mentis MJ, Dhawan V, Barnes A, et al. Effects of levodopa infusion on motor activation responses in Parkinson’s disease. Neurology 2002;59(2):220–6. Lozza C, Baron JC, Eidelberg D, Mentis MJ, Carbon M, Marie RM. Executive processes in Parkinson’s disease: FDG-PET and network analysis. Hum Brain Mapp 2004;22(3):236–45. Moeller JR, Nakamura T, Mentis MJ, Dhawan V, Spetsieres P, Antonini A, et al. Reproducibility of regional metabolic covariance patterns: comparison of four populations. J Nucl Med 1999;40(8):1264–9. Feigin A, Antonini A, Fukuda M, De Notaris R, Benti R, Pezzoli G, et al. Tc-99m ethylene cysteinate dimer SPECT in the differential diagnosis of parkinsonism. Mov Disord 2002;17(6):1265–70. Asanuma K, Tang C, Ma Y, Dhawan V, Mattis P, Edwards C, et al. Network modulation in the treatment of Parkinson’s disease. Brain 2006;129(Pt 10):2667–78. Eidelberg D, Moeller JR, Ishikawa T, Dhawan V, Spetsieris P, Chaly T, et al. Assessment of disease severity in parkinsonism with fluorine18-fluorodeoxyglucose and PET. J Nucl Med 1995;36(3):378–83. Moeller JR, Eidelberg D. Divergent expression of regional metabolic topographies in Parkinson’s disease and normal ageing. Brain 1997;120(Pt 12):2197–206. Asanuma K, Ma Y, Huang C, Carbon-Correll M, Edwards C, Raymond D, et al. The metabolic pathology of dopa-responsive dystonia. Ann Neurol 2005;57(4):596–600. Ma Y, Tang C, Spetsieres P, Dhawan V, Eidelberg D. Abnormal metabolic network activity in Parkinson’s disease: test–retest reproducibility. J Cereb Blood Flow Metab 2007;27(3):597–605. Eckert T, Feigin A, Lewis DE, Dhawan V, Frucht S, Eidelberg D. Regional metabolic changes in parkinsonian patients with normal dopaminergic imaging. Mov Disord 2007;22(2). Eckert T, Van Laere K, Tang C, Lewis DE, Edwards C, Santens P, et al. Quantification of PD-related network expression with ECD SPECT. Eur J Nucl Med Mol Imaging 2007;34(4):496–501. Feigin A, Leenders KL, Moeller JR, Missimer J, Kuenig G, Spetsieris P, et al. Metabolic network abnormalities in early Huntington’s disease: an [(18)F]FDG PET study. J Nucl Med 2001;42(11):1591–5. Feigin A, Budman C, Zgaljardic D, Dhawan V, Eidelberg D. Metabolic brain networks in Tourette syndrome. Mov Disord 2002;17(suppl 5):S339. Koeppe RA, Gilman S, Joshi A, Liu S, Little R, Junck L, et al. 11CDTBZ and 18F-FDG PET measures in differentiating dementias. J Nucl Med 2005;46(6):936–44. Scarmeas N, Habeck CG, Zarahn E, Anderson KE, Park A, Hilton J, et al. Covariance PET patterns in early Alzheimer’s disease and subjects with cognitive impairment but no dementia: utility in group discrimination and correlations with functional performance. Neuroimage 2004;23(1):35–45. Hauw JJ, Daniel SE, Dickson D, Horoupian DS, Jellinger K, Lantos PL, et al. Preliminary NINDS neuropathologic criteria for Steele– Richardson–Olszewski syndrome (progressive supranuclear palsy). Neurology 1994;44(11):2015–9. Wenning GK, Tison F, Ben-Shlomo Y, Daniel SE, Quinn N. Multiple system atrophy: a review of 203 pathologically proven cases. Mov Disord 1997;12(2):133–47. Eckert T, Sailer M, Kaufmann J, Schrader C, Peschel T, Bodammer N, et al. Differentiation of idiopathic Parkinson’s disease, multiple system atrophy, progressive supranuclear palsy, and healthy controls using magnetization transfer imaging. Neuroimage 2004;21(1):229–35. Seppi K, Schocke MF, Esterhammer R, Kremser C, Brenneis C, Mueller J, et al. Diffusion-weighted imaging discriminates progressive supranuclear palsy from PD, but not from the parkinson variant of multiple system atrophy. Neurology 2003;60(6):922–7.