The contribution of cerebellar proton magnetic resonance spectroscopy in the differential diagnosis among parkinsonian syndromes

Parkinsonism and Related Disorders 21 (2015) 929e937

Contents lists available at ScienceDirect

Parkinsonism and Related Disorders journal homepage: www.elsevier.com/locate/parkreldis

The contribution of cerebellar proton magnetic resonance spectroscopy in the differential diagnosis among parkinsonian syndromes Stefano Zanigni a, b, Claudia Testa a, b, Giovanna Calandra-Buonaura b, c, Luisa Sambati b, c, Maria Guarino d, Anna Gabellini c, e, Stefania Evangelisti a, b, Pietro Cortelli b, c, Raffaele Lodi a, b, *, Caterina Tonon a, b a

Functional MR Unit, Policlinico S. Orsola e Malpighi, Bologna (IT), Via Massarenti 9, 40138 Bologna, Italy Department of Biomedical and NeuroMotor Sciences, University of Bologna, Bologna (IT), Via U. Foscolo 7, 40123 Bologna, Italy c IRCCS Istituto Delle Scienze Neurologiche di Bologna, Bologna (IT), Via Altura 3, 40139 Bologna, Italy d Neurology Unit, Policlinico S. Orsola e Malpighi, Bologna (IT), Via Massarenti 9, 40138 Bologna, Italy e Neurology Unit, Ospedale Maggiore, Bologna (IT), Via B. Nigrisoli 2, 40133 Bologna, Italy b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 24 February 2015 Received in revised form 7 May 2015 Accepted 31 May 2015

Introduction: The in vivo differential diagnosis between idiopathic Parkinson's disease (PD) and atypical parkinsonian syndromes (PS), such as multiple system atrophy [MSA with a cerebellar (C) and parkinsonian (P) subtype] and progressive supranuclear palsy e Richardson's Syndrome (PSP-RS) is often challenging. Previous brain MR proton spectroscopy (1H-MRS) studies showed biochemical alterations in PS, despite results are conflicting. Cerebellum plays a central role in motor control and its alterations has been already demonstrated in atypical PS. The main aim of this study was to evaluate diagnostic accuracy of cerebellar 1H-MRS in the differential diagnosis between PD and atypical PS. Methods: We obtained 1H-MRS spectra from the left cerebellar hemisphere of 57 PS (21 PD, and 36 atypical PS) and 14 unaffected controls by using a 1.5 T GE scanner. N-acetyl-aspartate (NAA)/Creatine (Cr), choline-containing compounds (Cho)/Cr, myoinositol (mI)/Cr, and NAA/mI ratios were calculated. Results: NAA/Cr and NAA/mI ratios were significantly lower (p < 0.01) in atypical PS compared to PD and controls, and in MSA-C compared to PD, MSA-P, PSP-RS and controls. PSP-RS group showed reduced NAA/ Cr ratios compared to PD (p < 0.05) and controls (p < 0.05), and reduced NAA/mI compared to controls (p < 0.01). NAA/Cr ratio values higher than 1.016 showed 100% sensitivity and negative predictive value, 62% positive predictive value and 64% specificity in discriminating PD. Conclusion: Cerebellar biochemical alterations detected by using 1H-MRS could represent an adjunctive diagnostic tool to improve the differential diagnosis of PS. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Parkinson's disease MRS Cerebellum Progressive supranuclear palsy Multiple system atrophy

1. Introduction Parkinsonian syndromes (PS) are an heterogeneous group of neuropathologically distinct degenerative disorders characterized by extrapidamidal signs associated with pyramidal, cerebellar, autonomic and cognitive dysfunction. The main forms are represented by idiopathic Parkinson's disease (PD), Progressive

* Corresponding author. Functional MR Unit, Policlinico S. Orsola e Malpighi, Department of Biomedical and NeuroMotor Sciences (DiBiNeM), University of Bologna, Via Massarenti 9, 40138 Bologna (IT), Italy. E-mail address: [email protected] (R. Lodi). http://dx.doi.org/10.1016/j.parkreldis.2015.05.025 1353-8020/© 2015 Elsevier Ltd. All rights reserved.

Supranuclear Palsy e Richardson's Syndrome (PSP-RS), and the cerebellar and parkinsonian variant of Multiple System Atrophy (MSA-C and MSA-P respectively). Despite differences in clinical presentation, disease progression and response to dopaminergic treatment, the in vivo differential diagnosis remains challenging, especially in the earlier stages of the disease. International criteria have been established in order to permit a clinical “possible” or “probable” diagnosis, although the definite diagnosis is only anatomopathological and can be reached post-mortem [1e3]. In the last decades, various biomarkers have been evaluated in order to increase diagnostic accuracy of clinical criteria and to permit a differential diagnosis in the early disease's stages [4]. In particular, the

930

S. Zanigni et al. / Parkinsonism and Related Disorders 21 (2015) 929e937

matched to patients for age and sex, were selected among a sample of healthy volunteers, enrolled among University and Hospital workers and their relatives, that underwent brain MR in order to obtain normative values for quantitative MR parameters for clinical and research purposes. Demographic and clinical features of cases and controls are summarized in Table 2. Diagnosis were performed by neurologists with more than 10 years of experience in movement disorders (SZ, GC-B, MG, AG, PC) according to international criteria for PD [1], PSP [2], and MSA [3]. All PSP patients presented the Richardson's Syndrome variant (PSP-RS) [57]. Only 2 out of 21 PD patients were at Hoehn-Yahr stage 1 and presented parkinsonian signs only in one side (one patient right and the other left), while the other 19 presented bilateral symptoms with side prevalence, 9 with a right and 10 with a left prevalence. On the basis of clinical data obtained from the patients and from their clinical records at the day of MR scan, 12 PD patients fulfilled criteria for possible and 9 for probable PD, 1 PSP-RS patient was diagnosed as possible and 20 as probable PSP-RS, while all MSA patients fulfilled criteria for probable MSA. Because MR scans acquisitions were performed years before the analysis, in 7 PD patients the diagnosis evolved from possible to probable PD. The study protocol was approved by the local Ethical Committee and we obtained the written informed consent to personal data processing for research purposes from all participants.

contribute of neuroimaging in the differential diagnosis of PS has been established. The role of conventional brain MRI is crucial to distinguish secondary forms from neurodegenerative PS and to highlight pathognomonic alterations suggestive of different neurodegenerative syndromes such as the “hummingbird sign” due to tegmental midbrain atrophy in PSP and the “hot-cross bun sign” in MSA, signifying selective degeneration of ponto-cerebellar tracts [5e7]. Advanced MR techniques, such as volumetric and morphometric analysis, diffusion tensor imaging (DTI), proton MR spectroscopy (1H-MRS), and functional MRI, associated to conventional imaging, may contribute to define macro- and microstructural, biochemical and functional alterations in PS [5,8,9]. In particular, brain 1H-MRS permits an in vivo quantitative evaluation of biochemical profile within specific brain volumes of interest (VOIs). In particular, it has been shown that a reduction of N-acetylaspartate (NAA) concentration is related to neuronal and/or axonal damage as in neurodegenerative disorders such as dementias and ataxias, and increased myo-inositol (mI) is considered as a marker of glial reaction [9,10]. 1 H-MRS studies have been carried out in PS in order to better elucidate their pathophysiology and to contribute to differential diagnosis [10]. Different structures, such as substantia nigra, basal ganglia, hemispheric white and gray matter have been evaluated by this technique and the principal studies are summarized in Table 1. Because of the clinical variability of samples, different technical and methodological issues, the results of these studies are conflicting. Moreover, for the small size of the main brain structures involved in PD and atypical PS pathophysiology, such as substantia nigra and the basal ganglia, it is challenging to obtain an accurate metabolic profile by using 1H-MRS avoiding the partial volume effect of the adjacent structures [44]. The cerebellum is one major subcortical structure involved in motor control [45]. It is functionally and anatomically connected with the basal ganglia and its alterations have been demonstrated in PD [45]. Moreover, neuroimaging and neuropathological studies demonstrated macro- and microstructural changes in the cerebellum in MSA-C and PSP [47e50]. Previous studies evidenced the usefulness of cerebellar 1H-MRS in evaluating neurochemical alterations in inherited and sporadic ataxias [51e56]. Although the cerebellar involvement in movement disorders' pathophysiology, spectroscopic studies on cerebellum for the differential diagnosis of PS are lacking to date. In the only 1H-MRS study performed on cerebellar hemispheres Chaudhuri and colleagues reported normal NAA/Cr values in a small PD cohort compared to healthy controls [15]. The main aim of our study was to evaluate cerebellar biochemical profile in different forms of PS by using single voxel 1 H-MRS and to assess its accuracy in the differential diagnosis between PD and atypical PS. As secondary outcome, we also evaluated differences in cerebellar biochemical profile between parkinsonisms with cerebellar signs (cerebellar subgroup: PSP-RS and MSA-C) and without cerebellar signs (parkinsonian subgroup: PD and MSA-P).

Brain MR studies were performed using a 1.5 T GE® Medical Systems Signa HDx 15, equipped with a quadrature birdcage head coil. Structural imaging included axial FLAIR T2-weighted images (repetition time, TR ¼ 8000 ms, inversion time, TI ¼ 2000 ms, echo time, TE ¼ 93.5 ms, 3 mm slice thickness with no inter-slice gap), FSE coronal T2-weighted images (TR ¼ 7000 ms, TE ¼ 100 ms, 3 mm slice thickness), and 3D volumetric T1-weighted fast spoiled gradient-echo (FSPGR) images (TR ¼ 12.5 ms, TE ¼ 5.1 ms, TI ¼ 600 ms, 25.6 cm2 FOV; 1 mm isotropic voxels). MR images obtained from each subject were visualized by expert neuroradiologists (RL, CaT) in order to exclude secondary causes of parkinsonism and other signal intensity or morphology changes. A VOI of 2.0  2.0  1.5 cm3 was selected in the left cerebellar hemisphere (Fig. 1) using the three planes of high resolution 3D FSPGR T1 sequence to optimise the localisation. The selected VOI included cerebellar nuclei and white matter. Suppressed-water proton MR spectra were acquired using the PRESS single-voxel localization sequence (PROBE) with TR ¼ 4000 ms, TE ¼ 35 ms, and averaging 64 FIDs for each acquisition [58]. The acquisition time for MR spectroscopy sequence was 5 min 52 s. We decided to select the VOI in the left cerebellar hemisphere in order to standardize our acquisition and analysis protocol for all parkinsonian syndromes (with asymmetrical and symmetrical signs).

2. Materials and methods

2.3. MR spectroscopic data analysis

2.1. Subjects

Peak areas of NAA þ NAA glutamate (NAAG), Cr þ phospho-Cr (PCr), Glycero-phospho Cho (GPC)þ phospho Cho (PCh), and mI, were calculated using version 6.3 of the fitting program LCModel [59,60], a fully user-independent software, that analyzes spectra as a linear combination of complete model spectra of metabolite solutions in vitro. Metabolite content was expressed relative to Cr þ PCr. The exclusion criterion for metabolite evaluation was an LCModel estimated fitting error >20%, this being a reliable indicator of poor quality spectra.

We retrospectively included in this study all patients with PS referred between 2008 and 2013 to the Functional MR Unit, Policlinico S.Orsola e Malpighi, Bologna (IT), to perform brain MR as part of the diagnostic workup. We acquired and analyzed brain conventional MR images and 1H-MRS spectra of 57 PS: 21 PD, 36 atypical PS (15 MSA, 8 MSA-C and 7 MSA-P, and 21 PSP) and 14 controls, without evidence of neurological disorders. Controls,

2.2. Brain magnetic resonance imaging and spectroscopy acquisition

S. Zanigni et al. / Parkinsonism and Related Disorders 21 (2015) 929e937

931

Table 1 Summary of the 1H-MRS studies on parkinsonian syndromes. Sample

Magnetic field/1H-MRS TR/TE (ms) technique

VOI localisation

Abnormal findings

Bowen BC et al., 1995 [11]

14 PD 13 HC

1.5 T Single voxel

1500/270

Occipital lobe

[ Lac/NAA in PD

Davie CA et al., 1995 [12]

9 7 5 9

1.5 T Single voxel

2270/270

Lenticular nucleus

Y NAA/Cr and Y Cho/Cr in MSAP vs HC Y NAA/Cr in MSA-C vs HC

Davie CA et al., 1995 [13]

1.5 T 3 parkinsonism in ex-professional boxers STEAM Single voxel 6 PD 6 HC

2000/80

Lenticular nucleus (controlateral to the most affected body side)

Y NAA in ex-boxers vs HC

Holshouser BA et al., 1995 [14]

90 PD 61 HC

1.5 T Single voxel

1600/135

Nucleus striatum

Y NAA/Cho in an elderly subset or in untreated PD

Chaudhuri KR et al., 1996 [15]

10 PD with motor fluctuations 7 HC

1.5 T PRESS Single voxel

2000/136

Putamen (contralateral to the most affected body side) Cerebellar hemisphere

Y putaminal NAA/Cr and Y NAA/Cho in PD

Clarke CE et al., 1997 [16]

5 PD before and after apomorphine

1.5 T STEAM Single voxel

3000/20

Lenticular nucleus

e

Cruz CJ et al., 1997 [17]

10 PD 1 APS 13 HC

2T Single voxel

2000/272

Nucleus Striatum Thalamus

e

Davie CA et al., 1997 [18]

9 PSP 8 HC

1.5 T STEAM Single voxel

2270/270

Lenticular nucleus

Y NAA/Cr in PSP vs HC

Ellis CM et al., 1997 [19]

9 drug-naïve PD 7 PD with LID 11 HC

1.5 T PRESS Single voxel

Putamen

Y NAA/Cho in drug-naïve PD vs HC and PD with LID

Federico F et al., 1997 [20]

8 PD 5 PSP 9 HC

1.5 T Single voxel

1500/135

Lenticular nucleus

Y NAA/Cr and Y NAA/Cho in PSP vs HC

Federico F et al., 1997 [21]

12 PD 7 PSP 7 MSA 10 HC

1.5 T Single voxel

1500/135

Lenticular nucleus

Y NAA/Cho and Y NAA/Cr in PSP and MSA vs HC

Tedeschi G et al., 1997 [22]

10 PD 12 PSP 9 CBD 11 HC

1.5 T Single voxel

2200/272

Brainstem Caudatus Lenticular nucleus Thalamus Centrum semiovale Frontal, parietal, precentral, temporal and occipital cortices

Y NAA/Cr in brainstem, centrum semiovale and precentral cortices and Y NAA/ Cho in lenticular nucleus in PSP vs HC Y NAA/Cr in centrum semiovale and Y NAA/Cho in lenticular nucleus and parietal cortex in CBD vs HC

Choe BY et al., 1998 [23]

15 PD 10 HC

1.5 T Single voxel

2000/20

Substantia nigra Lenticular nucleus

Y NAA/Cr in the side contralateral to the most affected body side in PD

Hoang TQ et al., 1998 [24]

5 PD 15 HD 20 HC

1.5 T Single voxel

1500/30

Putamen Parietal white matter Occipital gray matter

e

Federico F et al., 1999 [25]

19 11 14 12

1.5 T Single voxel

1500/135

Lenticular nucleus

Y NAA/Cho in PSP and MSA vs PD and HC Y NAA/Cr in PD, PSP, MSA vs HC and in MSA vs PD

Hu MT et al., 1999 [26]

17 PD 10 HC

1.5 T 1500/130 Single and multi voxel

Temporoparietal and occipital cortex

Y NAA/Cr of temporo-parietal cortex in PD vs HC

TayloreRobinson SD et al., 1999 [27]

12 PD (7 with and 5 without LID) 12 HC

1.5 T Multi-voxel

Temporoparietal cortex

Y Cho/Cr of temporo-parietal cortex in PD vs HC Y NAA/Cr of temporo-parietal cortex in more severely affected PD vs HC

PD MSA-P MSA-C HC

PD PSP MSA HC

1500/130

(continued on next page)

932

S. Zanigni et al. / Parkinsonism and Related Disorders 21 (2015) 929e937

Table 1 (continued ) Sample

Magnetic field/1H-MRS TR/TE (ms) technique

VOI localisation

Abnormal findings

Abe K et al., 2000 [28]

23 12 18 19 10 20

1.5 T Single voxel

2000/135

Frontal lobe Putamen

Y NAA/Cr of frontal cortex in PSP, MSA, CBD and VP vs HC Y NAA/Cr of putamen in PSP, MSA, CBD and PD Y NAA/Cr of frontal cortex and putamen in CBD vs PD, MSA and VP Y NAA/Cr of putamen vs PD and MSA

Clarke CE et al., 2000 [29]

6 PD 6 MSA 6 HC

1.5 T STEAM Single voxel

3000/20

Lenticular nucleus

Y NAA/Cho and [ Cho/Cr in PD vs HC

Lucetti C et al., 2001 [30]

10 de-novo PD 10 HC

1.5 T PROBE Single voxel

2010/30

Y NAA/Cr of motor cortex in PD Medial surface motor cortex Striatum contralateral to the most affected body side

O'Neill J et al., 2002 [31]

10 PD 13 HC

1800/20 1.5 T STEAM Single and multi voxel

Camicioli RM et al., 2004 [32]

12 PD 10 HC

1.5 T PRESS Single voxel

2600/270,200,135,80,30 Posterior cingulate gyrus

Y NAA/Cr in PD vs HC

Watanabe H et al., 2004 [33]

24 MSA (13 MSA-C and 11 MSA-P) 11 PD 18 HC

3T Single voxel

2000/30

Putamen Pontine base Cerebral white matter (centrum semiovale)

Y NAA/Cr of pontine base in all MSA types and of putamen in MSA-P vs HC Y NAA/Cr of pontine base and putamen in MSA-P vs PD

Camicioli RM et al., 2007 [34]

44 PD 38 HC

1.5 T PRESS Single voxel

2600/80

Anterior cingulate gyrus Pre-SMA

Y NAA/Cr of pre-SMA in PD vs HC

Griffith HR et al., 2008 [35]

12 PD 12 HC

3T PRESS Single voxel

2000/80

Posterior cingulate gyrus

Y Glu/Cr in PD vs HC

Vasconcellos LF et al., 2009 [36] 12 PD 11 PSP 7 MSA-P 10 HC

1.5 T PRESS Single voxel

1500/50

Lenticular nucleus Midbrain Hippocampus White matter of frontal lobe

Y NAA/Cr of lenticular nucleus in PSP vs PD and HC Y NAA/Cr of the hippocampus in PSP vs HC Y NAA/Cho of the midbrain in PSP vs MSA-P and HC

Guevara CA et al., 2010 [37]

11 PD 13 PSP 11 MSA-P 6 MSA-C 18 HC

1.5 T 1500/35 Multi and single voxel

Lenticular nucleus Pallidum Putamen

Y NAA of putamen and pallidum in MSA-P and PSP vs PD and HC; Y NAA of pallidum in PSP vs MSA-P

Lehericy S et al., 2010 [38]

9 GP 10 PSP 9 HC

1.5 T PRESS Single voxel

1500/135

Lenticular nucleus

Y NAA/Cr in GP and PSP compared to HC

Takado Y et al., 2011 [39]

12 MSA-C 12 HC

3T PRESS Single voxel

1500/30

Pons and medulla

Y NAA/Cr and Cho/Cr in MSA-C vs HC [ mI/Cr and Cr in MSA-C vs HC

€ger A et al., 2013 [40] Gro

20 PD 5 PSP 3 MSA-P 2 CBD 22 HC

3T Multi-voxel

1350/30

Substantia nigra (rostral and caudal)

Rostral NAA/Cr < caudal in PD

€ger A et al., 2014 [41] Gro

21 PD 24 HC

3T Multi-voxel

1350/30

Substantia nigra (rostral and caudal)

[ NAA and Y Cho in rostral vs caudal region of SN in PD and HC Y NAA in caudal and rostral regions of SN in PD vs HC Rostral Cr > caudal in PD

Levin BE et al., 2014 [42]

12 PD 18 HC

3T Multi-voxel

1710/70

Gray and white matter in left hemispheric frontal, parietal, temporal and occipital lobes

Y NAA/Cr of temporal, occipital lobe, total brain gray matter and Y Cho/Cr in temporal lobe in PD vs HC

PD PSP MSA CBD VP HC

Substantia nigra Basal ganglia Frontal and parietal lobes

Y Cr in PD vs HC in all regions

S. Zanigni et al. / Parkinsonism and Related Disorders 21 (2015) 929e937

933

Table 1 (continued )

Zhou B et al., 2014 [43]

Sample

Magnetic field/1H-MRS TR/TE (ms) technique

VOI localisation

Abnormal findings

30 PD 20 HC

3T Multi-voxel

Substantia nigra

Y NAA/Cr, NAA/Cho, NAA/ (Cr þ Cho) in PD vs HC

3000/80

Legend: 1H-MRS: proton magnetic resonance spectroscopy; VOI: volume of interest; TR: Repetition Time; TE: Echo Time; PD: Idiopathic Parkinson's Disease; HC: Healthy Controls; Lac: Lactate; NAA: N-Acetyl-Aspartate; LID: levodopa induced dyskinesia; MSA: Multiple System Atrophy; MSA-P: Parkinsonian vaiant of MSA; MSA-C: Cerebellar variant; Cho: Choline; Cr: Creatine; PSP: Progressive Supranuclear Palsy; GP: Guadeloupean Parkinsonism; CBD: Cortico-Basal Degeneration; HD: Huntington's Disease; VP: Vascular Parkinsonism; pre-SMA: pre-supplementary motor area; APS: Atypical Parkinsonian Syndromes.

2.4. Statistical analysis Due to the small size of samples, we applied ManneWhitney test in order to compare metabolites' ratios between right- and leftpredominant PD groups. We tested the normal distribution of metabolites ratios by using ShapiroeWilk test. Therefore, in order to compare metabolites' ratios among PD, atypical PS and controls group we performed an one-way ANOVA followed by a Bonferroni post-hoc test for multiple comparisons. We applied the same method in a second analysis among PD, MSA-C, MSA-P, PSP and controls. In addition, we applied a t-test followed by Bonferroni correction for multiple comparisons to evaluate differences in metabolites' ratios between cerebellar and parkinsonian subgroups. We performed Pearson's correlation test in order to assess an association between age at evaluation and metabolites' ratios. We also determined specificity, sensitivity, positive predictive value (PPV), and negative predictive value (NPV) of NAA/Cr ratio in order to discriminate between PD and atypical PS. The cutoff value was determined by using receiver operating characteristics (ROC) curve analysis: the optimal value corresponded to the higher Youden's index (true positives e false positives rate). Statistical significance was set at p < 0.05. Analysis were performed by IBM® SPSS® v.21.

3. Results Mean age at evaluation in PSP-RS group resulted significantly higher than the other groups, although no differences in disease duration among groups were found. We found no significant correlation between age and metabolites' ratios (p ¼ 0.883). Disease stages distribution among patients with PD and atypical PS determined by using Hoehn and Yahr modified scale showed a modevalue of 3 in both groups (Table 2). Secondary causes of parkinsonism and other brain lesions were excluded in all subjects. We obtained spectra characterized by a mean signal-to-noise ratio (SNR) of 8.68 (±1.73) and mean Full Width Half Maximum (FWHM) of 4.26 Hz (±1.05). The estimated fitting error was below 20% in all spectra for all the metabolites of interest. Therefore, no acquired spectra were discarded for sub-optimal quality. We found no differences in left cerebellar hemispheric metabolites ratios in left or right predominant PD subgroups. In particular, we compared left-to right-predominant PD subgroups by using ManneWhitney test finding no significant differences in left cerebellar hemispheric metabolites ratios (right-predominant PD: NAA/ Cr ratio mean ± SD: 1.21 ± 0.16; NAA/mI ratio: 1.66 ± 0.22; leftpredominant PD: NAA/Cr ratio mean ± SD: 1.27 ± 0.22; NAA/mI ratio: 1.75 ± 0.27; NAA/Cr: p ¼ 0.605; NAA/mI: p ¼ 0.426). Moreover, we performed ANOVA test followed by Bonferroni post-hoc correction analysis among left-predominant PD, rightpredominant PD and atypical PS in order to evaluate the possible influence of side predominance on our results, in particular on the reduction of NAA/Cr and NAA/mI ratios in atypical PS compared to PD: we found significant differences in these ratios when

comparing left-predominant PD vs atypical PS, right-predominant PD vs atypical PS but not between PD subtypes (p < 0.01). Mean NAA/Cr and NAA/mI ratios in left cerebellar hemisphere were significantly lower in subjects with atypical PS compared to PD group (p < 0.01) and controls (p < 0.01), while PD subjects showed similar ratios to controls (Table 3 and Fig. 2). Moreover, mI/ Cr ratio was significantly higher in atypical PS compared to controls (p < 0.05). No significant differences were found in Cho/Cr ratio among PD, atypical PS and controls (Table 3). Mean NAA/Cr and NAA/mI ratios in MSA-P subjects were not dissimilar to PD and PSP-RS, while MSA-C group showed significantly lower values compared to PD (p < 0.01), MSA-P (p < 0.01), PSP-RS (p < 0.01) and controls (p < 0.01). MSA-C patients showed a significantly increased mI/Cr ratio compared to controls (p < 0.05). Moreover, in PSP-RS subjects, NAA/Cr ratio was significantly lower compared to PD (p < 0.01) and controls (p < 0.05), while NAA/mI ratio was significantly lower compared to controls (p < 0.01) (Table 3 and Fig. 2). We found no significant differences in Cho/Cr ratio among MSA, PSP-RS, PD and controls (Table 3). Cerebellar subgroup compared to parkinsonian showed significantly reduced NAA/Cr ratio (mean ± SD: 0.93 ± 0.21 and 1.21 ± 0.24, respectively) and NAA/mI ratio (1.31 ± 0.43 and 1.66 ± 0.31, respectively) in the left cerebellar hemisphere. ROC analysis showed that NAA/Cr ratio had the highest accuracy [area under the curve (AUC) ¼ 0.836; confidence interval (CI) ¼ 95%] in discriminating between PD and atypical PS (Fig. 3). Moreover, when comparing cerebellar versus parkinsonian subgroup, NAA/Cr ratio showed similar accuracy (AUC ¼ 0.842; CI ¼ 95%). Once determined the best cutoff value at 1.016 for both analysis, higher values showed 100% sensitivity and NPV, 62% PPV, and 64% specificity in discriminating PD from atypical PS, while 93% sensitivity, 72% specificity, 76% NPV and 91% PPV in differentiating parkinsonian vs. cerebellar subgroups.

4. Discussion Our study demonstrated that 1H-MRS is a sensitive technique able to detect the presence of biochemical alterations, such as reduced NAA and increased mI, markers of neuronal degeneration and glial reaction, respectively, in the cerebellar hemispheres in atypical PS. In particular, the biochemical profile that we found in MSA-C and PSP-RS patients, i.e. a marked NAA reduction in both groups and increased mI/Cr in MSA-C patients, was consistent with the pathological distribution of neurodegeneration in these two forms of parkinsonism [48,49]. Therefore, cerebellar metabolites quantification by using 1H-MRS represents an useful tool to differentiate PD and MSA-P from MSA-C and PSP-RS. In particular, NAA/Cr ratio was the best discriminant between PD and atypical PS, with high sensitivity and NPV. Although the age of patients in the PSP-RS group was higher than the other groups patients, we found neither significant differences in disease duration nor correlation between age at evaluation and NAA/Cr ratio, and therefore we can hypothesize that

934

S. Zanigni et al. / Parkinsonism and Related Disorders 21 (2015) 929e937

Table 2 Demographic and clinical features of cases and controls at the day of brain MR scan. Controls

N Age at evaluation (years) [median (range)] Sex [male/female] Disease duration (years) [median (range)] Hoehn-Yahr modified scale stage (percentual frequency)

14 64 (43e67) 8/6 Not applicable Not applicable

PD

21 63 (43e77) 16/5 3 (1e15) 1: 9.5% 1.5: 0% 2: 23.8% 2.5: 28.6% 3: 38.1% 4: 0% 5: 0%

Atypical PS All atypical PS

MSA

PSP-RS

C-subtype

P-subtype

36 67 (40e86) 22/14 4 (0.2e13) 1: 0% 1.5: 0% 2: 3.0% 2.5: 9.1% 3: 45.4% 4: 36.4% 5: 6.1%

8 57 (49e64) 4/4 6 (3e13)

7 61 (40e70) 6/1 3 (0.2e7)

21 73 (61e86) 12/9 4 (1e11)

Legend: PD: Idiopathic Parkinson's Disease; PS: Parkinsonian Syndromes; MSA: Multiple System Atrophy; PSP-RS: Progressive Supranuclear Palsy-Richardson's Syndrome.

Fig. 1. Typical Volume of Interest placement on the left cerebellar hemisphere and representative proton MR spectra from Idiopathic Parkinson's Disease (A), cerebellar variant of Multiple System Atrophy (B) and Progressive Supranuclear Palsy e Richardson's Syndrome (C) patients. Reduced N-Acetyl-Aspartate (NAA) peak is evident in B and C.

S. Zanigni et al. / Parkinsonism and Related Disorders 21 (2015) 929e937

935

Table 3 Proton MR spectroscopic cerebellar metabolites' ratios in parkinsonian syndromes and controls. Metabolites ratio

NAA/Cr Cho/Cr mI/Cr NAA/mI

Controls

1.25 0.27 0.63 1.96

(±0.24) (±0.04) (±0.12) (±0.40)

PD

1.24 0.27 0.73 1.71

Atypical PS

(±0.19) (±0.04) (±0.11) (±0.25)

All atypical PS

MSA

PSP-RS

C-subtype

P-subtype

0.97a,b (±0.25) 0.25 (±0.05) 0.74d (±0.14) 1.35a,b (±0.43)

0.70c (±0.13) 0.23 (±0.04) 0.84d (±0.11) 0.81c (±0.23)

1.13 0.28 0.75 1.51

(±0.36) (±0.05) (±0.13) (±0.43)

1.01a,e (±0.17) 0.25 (±0.06) 0.70 (±0.14) 1.49b (±0.32)

Legend: metabolites ratios are expressed as mean ± standard deviation; NAA: N-Acetyl-Aspartate; Cho: Choline; mI: myo-Inositol; Cr: Creatine; PS: parkinsonian syndromes; PD: Idiopatic Parkinson's Disease; MSA: Multiple System Atrophy; C: cerebellar; P: parkinsonian; PSP-RS: Progressive Supranuclear Palsy-Richardson's Syndrome. a Significantly decreased vs PD (p < 0.01). b Significantly decreased vs controls (p < 0.01). c Significantly decreased vs PD, PSP-RS, MSA-P and controls (p < 0.01). d Significantly increased vs controls (p < 0.05). e Significantly decreased vs controls (p < 0.05).

NAA/Cr reduction is related to the disease itself and not by age. When comparing PD with MSA-P vs PSP-RS with MSA-C, we found that the best cut-off value was the same as for the discrimination between PD and atypical PS, i.e. 1.016, with a similar diagnostic accuracy (AUC ¼ 0.842), slightly lower sensitivity (93%) and NPV (76%), and slightly higher PPV (91%) and specificity (72%). Therefore, we could argue that the inclusion of MSA-P sample in our study could affect the diagnostic properties of cerebellar MRS. Nevertheless, in a clinical perspective, NAA/Cr ratio values lower than 1.016 could indicate either atypical PS (PSP-RS or MSA-C or MSA-P) as well cerebellar subgroup (PSP-RS or MSA-C) with different degrees of sensitivity, specificity, NPV and PPV with the same diagnostic accuracy. Biochemical cerebellar alterations of PS patients that we found are in line with earlier neuroimaging and neuropathological studies demonstrating brainstem and basal ganglia alterations, although cerebellar metabolic changes in PD and PSP have been poorly explored to date. Previous spectroscopic studies of different forms of sporadic and hereditary ataxias, such as spinocerebellar ataxias, demonstrated an altered cerebellar biochemical profile characterized by reduced NAA/Cr ratios in MSA-C patients [51e54] and some authors found a correlation between the degree of NAA alterations and disease severity [52,54]. Moreover, a significant reduction of NAA in the basal ganglia (putamen and pallidum nuclei) and brainstem in atypical PS (MSA and PSP) compared to PD and controls has been demonstrated (Table 1). Watanabe et al. evidenced a reduction of NAA/Cr in MSA-

C and MSA-P groups in pontine base and in putamen in MSA-P patients in the earlier stages of the disease [33]. Previous studies found lower NAA/Cr, NAA/Cho and rostral to caudal NAA/Cr ratios in the SN in PD patients compared to atypical PS and controls [41,44]. Other studies instead failed to demonstrate alterations of the spectroscopic profile in basal ganglia [9,61] and cerebellar hemispheres in PD [15] compared to healthy controls. Morphological and microstructural alterations of cerebellum have been reported in both PSP and MSA. Conventional imaging and morphometric studies reported reduced size of MCP and SCP in MSA and PSP, respectively [46,62]. DTI studies demonstrated microstructural cerebellar alterations in MSA-C and MSA-P patients (MCP and cerebellar hemispheres) and in PSP (SCP and vermis) by showing increased tissue diffusivity [47,63,64]. In particular, in a previous study, we evidenced an increased median MD in the infratentorial compartment in MSA-P and eC patients compared to PD, PSP and healthy controls. Moreover, we found higher median MD values in PSP group compared to PD in the whole infratentorial compartment, in the brainstem and cerebellar vermis [50]. Overall, all these results are consistent with anatomopathological findings in atypical PS. The neuropathology of PSP is primarily characterized by the deposition of neurofibrillary tangles composed by Tau-protein within the central nervous system (CNS) with predominant damage in the basal ganglia and brainstem, associated with different degrees of midbrain and cortical atrophy, and degeneration of dentate nucleus and superior cerebellar peduncles [48]. Synucleinopathies, such as PD and MSA, are instead

Fig. 2. Box-plot graphics showing comparison of cerebellar N-Acetyl-Aspartate/Creatine (NAA/Cr) ratio; A: among idiopathic Parkinson's disease (PD), atypical parkinsonian syndromes (PS) and unaffected controls; B: among cerebellar (MSA-C) and parkinsonian (MSA-P) variant of Multiple System Atrophy, Progressive Supranuclear Palsy e Richardson's Syndrome (PSP-RS), PD and unaffected controls.

936

S. Zanigni et al. / Parkinsonism and Related Disorders 21 (2015) 929e937

5. Conclusions Cerebellar biochemical alterations in atypical PS can be detected by using 1H-MRS that may represent a viable tool to improve the accuracy of the differential diagnosis, in particular among PD, MSAC and PSP-RS. Further prospective studies on PS at the earlier stages of the disease will be necessary in order to corroborate the clinical utility of this technique. Disclosures All authors approved the final version of the article. Stefano Zanigni, Claudia Testa, Giovanna Calandra-Buonaura, Luisa Sambati, Maria Guarino, Anna Gabellini, Stefania Evangelisti, Raffaele Lodi and Caterina Tonon report no disclosures. Pietro Cortelli has received honoraria for speaking engagements or consulting activities with Allergan Italia, Boehringer Ingelheim Italia, Chelsea Therapeutics, GlaxoSmithKline S.p.A, Lundbeck Italy, Merck Sharp & Dohme (Italia), Teva, UCB Pharma S.p.A, Chiesi Farmaceutici, AbbVie srl. Fig. 3. ROC curve analysis of N-Acetyl-Aspartate/Creatine ratio in discriminating Parkinson's disease group from atypical parkinsonian syndromes group.

characterized by the deposition of a-synuclein within the CNS. In PD these aggregates are organized into neuronal Lewy bodies and Lewy neuritis, while in MSA a-synuclein deposits into oligodendrocytes cytoplasms [48]. The presence of neuronal loss in the substantia nigra associated with different degrees of Lewy-body deposition are the neuropathological hallmarks of PD, while MSA autoptic cases present different degrees of striato-nigral or olivoponto-cerebellar inclusions and degeneration depending on disease subtype (MSA-P or MSA-C respectively) [49,65]. Cerebellar metabolites' concentrations in our PD and MSA-P groups are instead not significantly dissimilar to those found in the control group, suggesting that cerebellar neurochemical involvement may be milder than in MSA-C and PSP-RS or not detectable with this technique. In particular, the absence of an altered cerebellar metabolic profile in the PD group is consistent with a previous study of Chaudhuri and colleagues [15]. Moreover, cerebellar biochemical profile in MSA-P and PD patients could not represent the right target to be studied, because of the neuropathological and therefore clinical paucity of cerebellar involvement in these parkinsonian syndromes. The biochemical study of cerebellum by using 1H-MRS presents the following technical advantages: first, cerebellar hemispheres are anatomically well-defined structures that can be easily manually delimitated during a scanning session, increasing the reproducibility of the VOI selection. Second, compared to other smaller structures such as SN and the basal ganglia, that region permits an adequate balance between the dimension of the VOI and the avoidance of the contamination signal from the adjacent structures, in order to achieve better quality spectra and higher Signal-toNoise Ratio (SNR). For these reasons, in our sample, no spectra were discarded because of poor quality. Third the additional scan time to a conventional MRI protocol is only 5 min 52 s. Because of the limited amount of samples, it has not been possible to stratify cases for disease duration, clinical severity and disease stage. To corroborate our findings it would be useful to test this diagnostic technique in a new sample of subjects with parkinsonian syndromes. Nevertheless, this study is the first evaluating cerebellar 1H-MRS profile and its usefulness in the differential diagnosis among PS.

Acknowledgments We thank the Technical and Nurses Staff of the MR Unit of Policlinico S. Orsola e Malpighi, Bologna, and its Coordinators, for the contribute in MR sequences acquisition. We also thank Claudio Bianchini and Francesco Campagna for the contribute in MR data analysis. References [1] D.J. Gelb, E. Oliver, S. Gilman, Diagnostic criteria for Parkinson disease, Arch. Neurol. 56 (1999) 33e39. [2] I. Litvan, Y. Agid, J. Jankovic, C. Goetz, J.P. Brandel, E.C. Lai, et al., Accuracy of clinical criteria for the diagnosis of progressive supranuclear palsy (SteeleRichardson-Olszewski syndrome), Neurology 46 (1996) 922e930. [3] S. Gilman, G.K. Wenning, P.A. Low, D.J. Brooks, C.J. Mathias, J.Q. Trojanowski, et al., Second consensus statement on the diagnosis of multiple system atrophy, Neurology 71 (2008) 670e676. [4] S. Sharma, C.S. Moon, A. Khogali, A. Haidous, A. Chabenne, C. Ojo, et al., Biomarkers in Parkinson's disease (recent update), Neurochem. Int. 63 (2013) 201e229. [5] P. Mahlknecht, A. Hotter, A. Hussl, R. Esterhammer, M. Schocke, K. Seppi, Significance of MRI in diagnosis and differential diagnosis of Parkinson's disease, Neurodegener. Dis. 7 (2010) 300e318. [6] D.J. Brooks, Imaging approaches to Parkinson disease, J. Nucl. Med. 51 (2010) 596e609. [7] L.A. Massey, C. Micallef, D.C. Paviour, S.S. O'Sullivan, H. Ling, D.R. Williams, et al., Conventional magnetic resonance imaging in confirmed progressive supranuclear palsy and multiple system atrophy, Mov. Disord. 27 (2012) 1754e1762. [8] C.W. Hess, E. Ofori, U. Akbar, M.S. Okun, D.E. Vaillancourt, The evolving role of diffusion magnetic resonance imaging in movement disorders, Curr. Neurol. Neurosci. Rep. 13 (2013) 400. [9] M. Rango, A. Arighi, P. Biondetti, B. Barberis, C. Bonifati, F. Blandini, et al., Magnetic resonance spectroscopy in Parkinson's disease and parkinsonian syndromes, Funct. Neurol. 22 (2007) 75e79. [10] G. Oz, J.R. Alger, P.B. Barker, R. Bartha, A. Bizzi, C. Boesch, et al., Clinical proton MR spectroscopy in central nervous system disorders, Radiology 270 (2014) 658e679. [11] B.C. Bowen, R.E. Block, J. Sanchez-Ramos, P.M. Pattany, D.A. Lampman, J.B. Murdoch, et al., Proton MR spectroscopy of the brain in 14 patients with Parkinson disease, Am. J. Neuroradiol. 16 (1995) 61e68. [12] C.A. Davie, G.K. Wenning, G.J. Barker, P.S. Tofts, B.E. Kendall, N. Quinn, et al., Differentiation of multiple system atrophy from idiopathic Parkinson's disease using proton magnetic resonance spectroscopy, Ann. Neurol. 37 (1995) 204e210. [13] C.A. Davie, Z. Pirtosek, G.J. Barker, D.P. Kingsley, P.H. Miller, A.J. Lees, Magnetic resonance spectroscopic study of parkinsonism related to boxing, J. Neurol. Neurosurg. Psychiatry 58 (1995) 688e691. €ller, J. Zijlmans, H. Kolem, D.B. Hinshaw Jr., [14] B.A. Holshouser, M. Komu, H.E. Mo et al., Localized proton NMR spectroscopy in the striatum of patients with idiopathic Parkinson's disease: a multicenter pilot study, Magn. Reson Med. 33 (1995) 589e594. [15] K.R. Chaudhuri, G.M. Lemmens, S.C. Williams, C. Ellis, C.M. Lloyd, J. Dawson, et

S. Zanigni et al. / Parkinsonism and Related Disorders 21 (2015) 929e937

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

[34]

[35] [36]

[37]

[38]

[39]

[40]

al., Proton magnetic resonance spectroscopy of the striatum in Parkinson's disease patients with motor response fluctuations, Park. Relat. Disord. 2 (1996) 63e67. C.E. Clarke, M. Lowry, A. Horsman, Unchanged basal ganglia N-acetylaspartate and glutamate in idiopathic Parkinson's disease measured by proton magnetic resonance spectroscopy, Mov. Disord. 12 (1997) 297e301. C.J. Cruz, M.J. Aminoff, D.J. Meyerhoff, S.H. Graham, M.W. Weiner, Proton MR spectroscopic imaging of the striatum in Parkinson's disease, Magn. Reson Imaging 15 (1997) 619e624. C.A. Davie, G.J. Barker, C. Machado, D.H. Miller, A.J. Lees, Proton magnetic resonance spectroscopy in Steele-Richardson-Olszewski syndrome, Mov. Disord. 12 (1997) 767e771. C.M. Ellis, G. Lemmens, S.C. Williams, A. Simmons, J. Dawson, P.N. Leigh, et al., Changes in putamen N-acetylaspartate and choline ratios in untreated and levodopa-treated Parkinson's disease: a proton magnetic resonance spectroscopy study, Neurology 49 (1997) 438e444. F. Federico, I.L. Simone, V. Lucivero, et al., Proton magnetic resonance spectroscopy in Parkinson's disease and progressive supranuclear palsy, J. Neurol. Neurosurg. Psychiatry 62 (1997) 239e242. F. Federico, I.L. Simone, V. Lucivero, M. De Mari, P. Giannini, G. Iliceto, et al., Proton magnetic resonance spectroscopy in Parkinson's disease and atypical parkinsonian disorders, Mov. Disord. 12 (1997) 903e909. G. Tedeschi, I. Litvan, S. Bonavita, A. Bertolino, N. Lundbom, N.J. Patronas, et al., Proton magnetic resonance spectroscopic imaging in progressive supranuclear palsy, Parkinson's disease and corticobasal degeneration, Brain 120 (1997) 1541e1552. B.Y. Choe, J.W. Park, K.S. Lee, B.C. Son, M.C. Kim, B.S. Kim, et al., Neuronal laterality in Parkinson's disease with unilateral symptom by in vivo 1H magnetic resonance spectroscopy, Invest Radiol. 33 (1998) 450e455. T.Q. Hoang, S. Bluml, D.J. Dubowitz, R. Moats, O. Kopyov, D. Jacques, et al., Quantitative proton-decoupled 31P MRS and 1H MRS in the evaluation of Huntington's and Parkinson's diseases, Neurology 50 (1998) 1033e1040. F. Federico, I.L. Simone, V. Lucivero, D.M. Mezzapesa, M. de Mari, P. Lamberti, et al., Usefulness of proton magnetic resonance spectroscopy in differentiating parkinsonian syndromes, Ital. J. Neurol. Sci. 20 (1999) 223e229. M.T. Hu, S.D. Taylor-Robinson, K.R. Chaudhuri, J.D. Bell, R.G. Morris, C. Clough, et al., Evidence for cortical dysfunction in clinically non-demented patients with Parkinson's disease: a proton MR spectroscopy study, J. Neurol. Neurosurg. Psychiatry 67 (1999) 20e26. S.D. Taylor-Robinson, N. Turjanski, S. Bhattacharya, J.P. Seery, J. Sargentoni, D.J. Brooks, et al., A proton magnetic resonance spectroscopy study of the striatum and cerebral cortex in Parkinson's disease, Metab. Brain Dis. 14 (1999) 45e55. K. Abe, H. Terakawa, M. Takanashi, Y. Watanabe, H. Tanaka, N. Fujita, et al., Proton magnetic resonance spectroscopy of patients with parkinsonism, Brain Res. Bull. 52 (2000) 589e595. C.E. Clarke, M. Lowry, Basal ganglia metabolite concentrations in idiopathic Parkinson's disease and multiple system atrophy measured by proton magnetic resonance spectroscopy, Eur. J. Neurol. 7 (2000) 661e665. C. Lucetti, P. Del Dotto, G. Gambaccini, S. Bernardini, M.C. Bianchi, M. Tosetti, et al., Proton magnetic resonance spectroscopy (1H-MRS) of motor cortex and basal ganglia in de novo Parkinson's disease patients, Neurol. Sci. 22 (2001) 69e70. J. O'Neill, N. Schuff, W.J. Marks Jr., R. Feiwell, M.J. Aminoff, M.W. Weiner, Quantitative 1H magnetic resonance spectroscopy and MRI of Parkinson's disease, Mov. Disord. 17 (2002) 917e927. R.M. Camicioli, J.R. Korzan, S.L. Foster, N.J. Fisher, D.J. Emery, A.C. Bastos, et al., Posterior cingulate metabolic changes occur in Parkinson's disease patients without dementia, Neurosci. Lett. 354 (2004) 177e180. H. Watanabe, H. Fukatsu, M. Katsuno, M. Sugiura, K. Hamada, Y. Okada, et al., Multiple regional 1H-MR spectroscopy in multiple system atrophy: NAA/Cr reduction in pontine base as a valuable diagnostic marker, J. Neurol. Neurosurg. Psychiatry 75 (2004) 103e109. R.M. Camicioli, C.C. Hanstock, T.P. Bouchard, M. Gee, N.J. Fisher, W.R. Martin, Magnetic resonance spectroscopic evidence for presupplementary motor area neuronal dysfunction in Parkinson's disease, Mov. Disord. 22 (2007) 382e386. H.R. Griffith, O.C. Okonkwo, T. O'Brien, J.A. Hollander, Reduced brain glutamate in patients with Parkinson's disease, NMR Biomed. 21 (2008) 381e387. L.F. Vasconcellos, S.A. Novis, D.M. Moreira, A.L. Rosso, A.C. Leite, Neuroimaging in Parkinsonism: a study with magnetic resonance and spectroscopy as tools in the differential diagnosis, Arq. Neuropsiquiatr. 67 (2009) 1e6. C.A. Guevara, C.R. Blain, D. Stahl, D.J. Lythgoe, P.N. Leigh, G.J. Barker, Quantitative magnetic resonance spectroscopic imaging in Parkinson's disease, progressive supranuclear palsy and multiple system atrophy, Eur. J. Neurol. 17 (2010) 1193e1202. ricy, A. Hartmann, A. Lannuzel, D. Galanaud, C. Delmaire, S. Lehe e, et al., Magnetic resonance imaging lesion pattern in GuadeM.J. Bienaime loupean parkinsonism is distinct from progressive supranuclear palsy, Brain 133 (2010) 2410e2425. Y. Takado, H. Igarashi, K. Terajima, T. Shimohata, T. Ozawa, K. Okamoto, et al., Brainstem metabolites in multiple system atrophy of cerebellar type: 3.0-T magnetic resonance spectroscopy study, Mov. Disord. 26 (2011) 1297e1302. €ger, B. Bender, I. Wurster, G.L. Chadzynski, U. Klose, D. Berg, A. Gro

[41]

[42]

[43]

[44]

[45] [46]

[47]

[48]

[49]

[50]

[51]

[52]

[53] [54]

[55]

[56]

[57]

[58]

[59] [60] [61]

[62]

[63]

[64]

[65]

937

Differentiation between idiopathic and atypical parkinsonian syndromes using three-dimensional magnetic resonance spectroscopic imaging, J. Neurol. Neurosurg. Psychiatry 84 (2013) 644e649. €ger, R. Kolb, R. Scha €fer, U. Klose, Dopamine reduction in the substantia A. Gro nigra of Parkinson's disease patients confirmed by in vivo magnetic resonance spectroscopic imaging, PLoS One 9 (2014) e84081. B.E. Levin, H.L. Katzen, A. Maudsley, J. Post, C. Myerson, V. Govind, et al., Whole-brain proton MR spectroscopic imaging in Parkinson's disease, J. Neuroimaging 24 (2014) 39e44. B. Zhou, F. Yuan, Z. He, C. Tan, Application of proton magnetic resonance spectroscopy on substantia nigra metabolites in Parkinson's disease, Brain Imaging Behav. 8 (2014) 97e101. ricy, M.A. Sharman, C.L. Dos Santos, R. Paquin, C. Gallea, Magnetic S. Lehe resonance imaging of the substantia nigra in Parkinson's disease, Mov. Disord. 27 (2012) 822e830. T. Wu, M. Hallett, The cerebellum in Parkinson's disease, Brain 136 (2013) 696e709. A. Quattrone, G. Nicoletti, D. Messina, F. Fera, F. Condino, P. Pugliese, et al., MR imaging index for differentiation of progressive supranuclear palsy from Parkinson disease and the Parkinson variant of multiple system atrophy, Radiology 246 (2008) 214e221. G. Rizzo, P. Martinelli, D. Manners, C. Scaglione, C. Tonon, P. Cortelli, et al., Diffusion-weighted brain imaging study of patients with clinical diagnosis of corticobasal degeneration, progressive supranuclear palsy and Parkinson's disease, Brain 131 (2008) 2690e2700. D.W. Dickson, Z. Ahmed, A.A. Algom, Y. Tsuboi, K.A. Josephs, Neuropathology of variants of progressive supranuclear palsy, Curr. Opin. Neurol. 23 (2010) 394e400. G.M. Halliday, J.L. Holton, T. Revesz, D.W. Dickson, Neuropathology underlying clinical variability in patients with synucleinopathies, Acta Neuropathol. 122 (2011) 187e204. G. Nicoletti, G. Rizzo, G. Barbagallo, C. Tonon, F. Condino, D. Manners, et al., Diffusivity of cerebellar hemispheres enables discrimination of cerebellar or parkinsonian multiple system atrophy from progressive supranuclear palsyRichardson syndrome and Parkinson disease, Radiology 267 (2013) 843e850. N. Ikuta, Proton magnetic resonance spectroscopy and single photon emission CT in patients with olivopontocerebellar atrophy, Rinsho Shinkeigaku 38 (1998) 289e294. M. Mascalchi, M. Cosottini, F. Lolli, F. Salvi, C. Tessa, M. Macucci, et al., Proton MR spectroscopy of the cerebellum and pons in patients with degenerative ataxia, Radiology 223 (2002) 371e378. M. Viau, Y. Boulanger, Characterization of ataxias with magnetic resonance imaging and spectroscopy, Park. Relat. Disord. 10 (2004) 335e351. S.M. Boesch, C. Wolf, K. Seppi, S. Felber, G.K. Wenning, M. Schocke, Differentiation of SCA2 from MSA-C using proton magnetic resonance spectroscopic imaging, J. Magn. Reson Imaging 25 (2007) 564e569. G. Oz, I. Iltis, D. Hutter, W. Thomas, K.O. Bushara, C.M. Gomez, Distinct neurochemical profiles of spinocerebellar ataxias 1, 2, 6, and cerebellar multiple system atrophy, Cerebellum 10 (2011) 208e217. J.F. Lirng, P.S. Wang, H.C. Chen, B.W. Soong, W.Y. Guo, H.M. Wu, et al., Differences between spinocerebellar ataxias and multiple system atrophycerebellar type on proton magnetic resonance spectroscopy, PLoS One 7 (2012) e47925. D.R. Williams, R. de Silva, D.C. Paviour, A. Pittman, H.C. Watt, L. Kilford, et al., Characteristics of two distinct clinical phenotypes in pathologically proven progressive supranuclear palsy: Richardson's syndrome and PSP-parkinsonism, Brain 128 (2005) 1247e1258. R. Lodi, P. Parchi, C. Tonon, D. Manners, S. Capellari, R. Strammiello, et al., Magnetic resonance diagnostic markers in clinically sporadic prion disease: a combined brain magnetic resonance imaging and spectroscopy study, Brain 132 (2009) 2669e2679. S.W. Provencher, Estimation of metabolite concentrations from localized in vivo proton NMR spectra, Magn. Reson Med. 30 (1993) 672e679. S.W. Provencher, Automatic quantitation of localized in vivo 1H spectra with LCModel, NMR Biomed. 14 (2001) 260e264. N. Weiduschat, X. Mao, M.F. Beal, M.J. Nirenberg, D.C. Shungu, C. Henchcliffe, Usefulness of proton and phosphorus MR spectroscopic imaging for early diagnosis of Parkinson's disease, J. Neuroimaging 25 (2013) 105e110. G. Nicoletti, F. Fera, F. Condino, W. Auteri, O. Gallo, P. Pugliese, et al., MR imaging of middle cerebellar peduncle width: differentiation of multiple system atrophy from Parkinson disease, Radiology 239 (2006) 825e830. G. Nicoletti, R. Lodi, F. Condino, et al., Apparent diffusion coefficient measurements of the middle cerebellar peduncle differentiate the Parkinson variant of MSA from Parkinson's disease and progressive supranuclear palsy, Brain 129 (2006) 2679e2687. G. Nicoletti, R. Lodi, F. Condino, C. Tonon, F. Fera, Malucelli, et al., Apparent diffusion coefficient of the superior cerebellar peduncle differentiates progressive supranuclear palsy from Parkinson's disease, Mov. Disord. 23 (2008) 2370e2376. T. Ozawa, D. Paviour, N.P. Quinn, K.A. Josephs, H. Sangha, L. Kilford, et al., The spectrum of pathological involvement of the striatonigral and olivopontocerebellar systems in multiple system atrophy: clinicopathological correlations, Brain 127 (2004) 2657e2671.