Parkinsonism and Related Disorders 20 (2014) 1004e1008
Contents lists available at ScienceDirect
Parkinsonism and Related Disorders journal homepage: www.elsevier.com/locate/parkreldis
Reduced thalamic volume in Parkinson disease with REM sleep behavior disorder: Volumetric study M. Salsone b, 1, A. Cerasa c, 1, G. Arabia a, M. Morelli a, A. Gambardella a, L. Mumoli a, c, B. Vescio c, A. Quattrone a, c, * R. Nistico a b c
Institute of Neurology, Department of Medical Sciences, University Magna Graecia, Germaneto, Catanzaro, Italy Neuroimaging Research Unit, National Research Council, Catanzaro, Italy Neuroimaging Research Unit, Institute of Bioimaging and Molecular Physiology, National Research Council, Germaneto, Catanzaro, Italy
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 February 2014 Received in revised form 16 May 2014 Accepted 17 June 2014
Introduction: REM sleep behavior disorder (RBD) is a common non motor feature of Parkinson's Disease (PD) affecting about half the patients with this disease. Distinct structural brain tissue abnormalities have been reported in several regions modulating REM sleep of the patients with idiopathic RBD. At the present time, there are no conventional MRI studies investigating patients with PD associated with RBD. Methods: Herein, we used voxel-based morphometry (VBM) to detect the neuroanatomical profile of PD patients with and without RBD. Optimized VBM was applied to the MRI brain images in 11 PD patients with RBD (PD-RBD), 11 PD patients without RBD (PD) and 18 age-and sex-matched controls. To corroborate VBM findings we used automated volumetric method (FreeSurfer) to quantify subcortical brain regions volumes. Patients and controls also underwent DAT-SPECT and cardiac MIBG scintigraphies. Results: The VBM analysis showed markedly reduced gray matter volume in the right thalamus of PDRBD patients in comparison with PD patients and controls. Automatic thalamic segmentation in PD-RBD patients showed a bilaterally reduced thalamic volume as compared with PD patients or controls. All PD patients (with and without RBD) showed a reduced tracer uptake on DAT-SPECT and cardiac MIBG scintigraphies as compared to controls. Conclusions: Our findings suggest that the presence of RBD symptoms in PD patients is associated with a reduced thalamic volume suggesting a pathophysiologic role of the thalamus in the complex circuit causing RBD. © 2014 Published by Elsevier Ltd.
Keywords: Parkinson disease REM sleep behavior disorder Resonance imaging Voxel-based morphometry Automated segmentation method
1. Introduction Parkinson's disease (PD) is a progressive neurodegenerative disorder clinically characterized by a core motor features including, akinesia, rigidity and tremor. PD is also associated with non motor symptoms as hyposmia and several sleep disorders including REM sleep behavior disorder (RBD), a parasomnia characterized by dream-enacting behaviors and loss of atonia during the REM sleep. The prevalence of RBD symptoms in patients with PD based on clinical history varies from 15% to 46% while it is estimated in the range between 46% and 58% when the clinical diagnosis of RBD is confirmed using video-polysomnography. In patients with PD
* Corresponding author. Institute of Neurology, Department of Medical Sciences, University “Magna Graecia”, 88100 Germaneto, Catanzaro, Italy. Tel.: þ39 0961 3647075; fax: þ39 0961 3647177. E-mail address: [email protected] (A. Quattrone). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.parkreldis.2014.06.012 1353-8020/© 2014 Published by Elsevier Ltd.
exhibiting RBD this parasomnia precedes the onset of parkinsonism in 18e22% of the cases [1]. There is increasing evidence that the presence of RBD in PD patients identifies a clinical subtype of the disease. Indeed, in patients with PD, RBD is associated with older age, longer disease duration [1], rigid-akinetic form of PD and more severe parkinsonian symptoms [2]. These patients may also have increased autonomic dysfunction and higher risk to develop dementia and therefore worse prognosis [3]. Although the clinical phenotype of patients with PD associated with RBD is well known remain still uncertain why only a subset of PD patients develop RBD. Voxel-based morphometry (VBM) is a well-known unbiased quantitative MR imaging (MRI) method that provides an objective measure of gray and white matter volume changes across the entire brain. Recently, a VBM study [4] conducted in patients with idiopathic RBD (iRBD) showed a significant increase of gray matter density in both hippocampi whereas an other VBM study [5] revealed a significant gray matter reduction in the cerebellum,
M. Salsone et al. / Parkinsonism and Related Disorders 20 (2014) 1004e1008
tegmental portion of the pons and parahippocampal gyrus. Taken together, these findings suggest that many structures with different VBM pattern may be involved in the complex circuit causing RBD. Several studies have recently validated the use of automated volumetric methods such as FreeSurfer to quantify cortical and subcortical volumes. FreeSurfer calculates brain sub-volumes by assigning a neuroanatomical label to label to each voxel in an MRI volume on the basis of probabilistic information estimated automatically from a manually labeled training set [6]. To our knowledge, there are no conventional MRI reports investigating PD patients associated with RBD. Owing to this lack of investigation, the purpose of the current study was to detect the neuroanatomical profile of PD patients with and without RBD using combined VBM and FreeSurfer analyses. 2. Patients and methods A total of 11 patients with PD associated with RBD (PD-RBD) and 11 patients with PD without RBD (PD) were recruited from the Neurology Unit of the University “Magna Graecia” of Catanzaro. All subjects met the UK Parkinson's Disease Society Brain Bank clinical diagnostic criteria [7] for PD. DAT-SPECT imaging and cardiac MIBG scintigraphy were also performed in all patients to support a degenerative parkinsonian condition [8]. In all PD patients the RBD symptoms were assessed using the RBD Single-Question Screen [9] (RBD 1Q) a single “yes-no” question that queries the classic dream enactment behavior of RBD. We also performed in all PD patients (with and without RBD) an audio-visual polysomnography (PSG) to confirm the clinical diagnosis of RBD symptoms in PD-RBD group and exclude their presence in PD group. On PSG, a prominent muscle activity in REM sleep associated with abnormal behaviors was required to confirm the clinical diagnosis of RBD [10]. Exclusion criteria were: 1) current treatment with medications know modify REM sleep architecture and muscle tone as serotonin reuptake inhibitors; 2) evidence of structural abnormalities in the brain affecting the gray matter; 3)head movement artifacts during the MRI section; 4) evidence of global cognitive impairment (MMSE 26) [11]. Before MRI examination, patients underwent a careful clinical assessment including Hoehn-Yahr (HeY) staging [12] Unified Parkinson's Disease Rating scale (UPDRS) scores [13] and neuropsychological evaluation including MMSE, verbal fluency, token test, Frontal Assessment Battery (FAB) and Back Depression Inventory. Patients were compared to eighteen age-and sexmatched control subjects who had no history suggestive of RBD or other neurological or psychiatric diseases and showed normal MRI of the brain, both normal DAT-SPECT and cardiac MIBG scintigraphy.
3. Magnetic resonance imaging Brain MRI was performed according to our routine protocol by a 3 T scanner with an 8-channel head coil (Discovery MR-750, GE, Milwaukee, WI, USA). Structural MRI data were acquired using a 3D T1-weighted spoiled gradient echo (SPGR) sequence with the following parameters: TR: 3.7 ms, TE: 9.2 ms, flip angle 12 , voxelsize 1 1 1 mm3. Subjects were positioned to lie comfortably in the scanner with a forehead-restraining strap and various foam pads to ensure head fixation. 4. Voxel-based morphometry Data were processed and examined using the SPM8 software (Wellcome Trust Centre for Neuroimaging, London, UK), where we applied VBM implemented in the VBM8 toolbox (http://dbm.neuro. uni-jena.de/vbm.html) with default parameters incorporating the DARTEL toolbox that was used to obtain a high-dimensional normalization protocol. Images were bias-corrected, tissue classified, and registered using linear (12-parameter affine) and non-linear transformations (warping), within a unified model [14]. Subsequently, the warped gray matter (GM) segments were affine transformed into Montreal Neurological Institute (MNI) space and were scaled by the Jacobian determinants of the deformations (modulated GM volumes). Finally, the modulated volumes were smoothed with a Gaussian kernel of 8 mm full width at half maximum (FWHM). The GM volume maps were statistically analyzed using the general linear model based on Gaussian random field theory. Statistical analysis
1005
consisted of an analysis of covariance (AnCOVA) used for investigating the main effect of group (F-test). The advantage of an SPM-related Fstatistic is that increases and decreases of GM volumes are analyzed together to detect morphological changes in three or more groups. Age and total intracranial volume (ICV) were included in the model as covariates of no-interest. Two approaches to statistical threshold maps were applied. First, we applied a conservative approach with a whole-brain statistical threshold correction (p < 0.05, family-wise error (FWE)). Second, since that in vivo evidence of GM abnormalities in PD patients with RBD has never been reported, the data were also presented by using a less-stringent, uncorrected threshold (p < 0.001, cluster (k) threshold ¼ 10 voxels) to detect subtle morphological changes. 4.1. Automated subcortical volumetry To corroborate voxel-based findings we further performed automated labeling and quantification of subcortical brain regions volume using FreeSurfer 5.0. The automated procedures for volumetric measures of several deep gray matter structures have been previously described [6,15]. The automated subcortical segmentation performed by FreeSurfer required these steps: Firstly, an optimal linear transform is computed that maximizes the likelihood of the input image, given an atlas constructed from manually labeled images. A non-linear transform is then initialized with the linear one, and the image is allowed to further deform to better match the atlas. Finally, a Bayesian segmentation procedure is performed, and the maximum a posteriori estimate of the labeling is computed. This approach provides advantages similar to manual ROI drawing [6,16] without the potential for rater bias, offering an anatomically accurate rendering of regional volumes. ICV was Table 1 Demographic and clinical characteristics and neuroimaging features of PD patients and controls. Variables
PD-RBD (n ¼ 11)
Age (y) Sex (m/w) Age at onset (y) Disease duration (y) Hoehn and Yahr (HY) stage UPDRS UPDRS-ME Levodopa dosage (mg/die) Rigid-akinetic form (n) Onset of RBD symptoms (y) Cognitive data MMSE Verbal fluency Token test F.A.B. Beck depression inventory DAT-SPECT Put/Cau right Put/Cau left Cardiac MIBG scintigraphy H/M ratio of delayed images
66.6 8/3 61.9 4.72 1.95
± 7.4 ± 6.4 ± 4.07 ± 0.57
PD (n ¼ 11) 66.9 8/3 62.5 4.36 1.86
± 7.9 ± 7.27 ± 4.2 ± 0.59
Controls (n ¼ 18)
P-value
65.1 ± 7.8 13/5 e e e
0.79a
0.85b 0.70b 0.92c
30 ± 15.4 21.6 ± 10.4 563.6 ± 167.3
28.9 ± 10.9 19.9 ± 10.3 547.7 ± 155.4
e e e
3/11
2/11
e
0.83b 0.69c 0.64c
2.95 ± 1.09
27.4 24.4 30.4 13.6 7.45
± ± ± ± ±
2.87 8.77 5.03 1.98 4.01
26.9 24.6 29.5 14.3 7.54
± ± ± ± ±
3.27 12.0 2.08 2.36 3.72
27.9 27.2 31.0 15.2 6.50
± ± ± ± ±
1.85 7.38 1.88 1.74 6.89
0.86d 0.65a 0.14d 0.09a 0.31d
1.70 ± 0.22 1.73 ± 0.19
1.66 ± 0.21 1.64 ± 0.19
2.46 ± 0.09 2.45 ± 0.11
Contents lists available at ScienceDirect
Parkinsonism and Related Disorders journal homepage: www.elsevier.com/locate/parkreldis
Reduced thalamic volume in Parkinson disease with REM sleep behavior disorder: Volumetric study M. Salsone b, 1, A. Cerasa c, 1, G. Arabia a, M. Morelli a, A. Gambardella a, L. Mumoli a, c, B. Vescio c, A. Quattrone a, c, * R. Nistico a b c
Institute of Neurology, Department of Medical Sciences, University Magna Graecia, Germaneto, Catanzaro, Italy Neuroimaging Research Unit, National Research Council, Catanzaro, Italy Neuroimaging Research Unit, Institute of Bioimaging and Molecular Physiology, National Research Council, Germaneto, Catanzaro, Italy
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 February 2014 Received in revised form 16 May 2014 Accepted 17 June 2014
Introduction: REM sleep behavior disorder (RBD) is a common non motor feature of Parkinson's Disease (PD) affecting about half the patients with this disease. Distinct structural brain tissue abnormalities have been reported in several regions modulating REM sleep of the patients with idiopathic RBD. At the present time, there are no conventional MRI studies investigating patients with PD associated with RBD. Methods: Herein, we used voxel-based morphometry (VBM) to detect the neuroanatomical profile of PD patients with and without RBD. Optimized VBM was applied to the MRI brain images in 11 PD patients with RBD (PD-RBD), 11 PD patients without RBD (PD) and 18 age-and sex-matched controls. To corroborate VBM findings we used automated volumetric method (FreeSurfer) to quantify subcortical brain regions volumes. Patients and controls also underwent DAT-SPECT and cardiac MIBG scintigraphies. Results: The VBM analysis showed markedly reduced gray matter volume in the right thalamus of PDRBD patients in comparison with PD patients and controls. Automatic thalamic segmentation in PD-RBD patients showed a bilaterally reduced thalamic volume as compared with PD patients or controls. All PD patients (with and without RBD) showed a reduced tracer uptake on DAT-SPECT and cardiac MIBG scintigraphies as compared to controls. Conclusions: Our findings suggest that the presence of RBD symptoms in PD patients is associated with a reduced thalamic volume suggesting a pathophysiologic role of the thalamus in the complex circuit causing RBD. © 2014 Published by Elsevier Ltd.
Keywords: Parkinson disease REM sleep behavior disorder Resonance imaging Voxel-based morphometry Automated segmentation method
1. Introduction Parkinson's disease (PD) is a progressive neurodegenerative disorder clinically characterized by a core motor features including, akinesia, rigidity and tremor. PD is also associated with non motor symptoms as hyposmia and several sleep disorders including REM sleep behavior disorder (RBD), a parasomnia characterized by dream-enacting behaviors and loss of atonia during the REM sleep. The prevalence of RBD symptoms in patients with PD based on clinical history varies from 15% to 46% while it is estimated in the range between 46% and 58% when the clinical diagnosis of RBD is confirmed using video-polysomnography. In patients with PD
* Corresponding author. Institute of Neurology, Department of Medical Sciences, University “Magna Graecia”, 88100 Germaneto, Catanzaro, Italy. Tel.: þ39 0961 3647075; fax: þ39 0961 3647177. E-mail address: [email protected] (A. Quattrone). 1 These authors contributed equally to this work. http://dx.doi.org/10.1016/j.parkreldis.2014.06.012 1353-8020/© 2014 Published by Elsevier Ltd.
exhibiting RBD this parasomnia precedes the onset of parkinsonism in 18e22% of the cases [1]. There is increasing evidence that the presence of RBD in PD patients identifies a clinical subtype of the disease. Indeed, in patients with PD, RBD is associated with older age, longer disease duration [1], rigid-akinetic form of PD and more severe parkinsonian symptoms [2]. These patients may also have increased autonomic dysfunction and higher risk to develop dementia and therefore worse prognosis [3]. Although the clinical phenotype of patients with PD associated with RBD is well known remain still uncertain why only a subset of PD patients develop RBD. Voxel-based morphometry (VBM) is a well-known unbiased quantitative MR imaging (MRI) method that provides an objective measure of gray and white matter volume changes across the entire brain. Recently, a VBM study [4] conducted in patients with idiopathic RBD (iRBD) showed a significant increase of gray matter density in both hippocampi whereas an other VBM study [5] revealed a significant gray matter reduction in the cerebellum,
M. Salsone et al. / Parkinsonism and Related Disorders 20 (2014) 1004e1008
tegmental portion of the pons and parahippocampal gyrus. Taken together, these findings suggest that many structures with different VBM pattern may be involved in the complex circuit causing RBD. Several studies have recently validated the use of automated volumetric methods such as FreeSurfer to quantify cortical and subcortical volumes. FreeSurfer calculates brain sub-volumes by assigning a neuroanatomical label to label to each voxel in an MRI volume on the basis of probabilistic information estimated automatically from a manually labeled training set [6]. To our knowledge, there are no conventional MRI reports investigating PD patients associated with RBD. Owing to this lack of investigation, the purpose of the current study was to detect the neuroanatomical profile of PD patients with and without RBD using combined VBM and FreeSurfer analyses. 2. Patients and methods A total of 11 patients with PD associated with RBD (PD-RBD) and 11 patients with PD without RBD (PD) were recruited from the Neurology Unit of the University “Magna Graecia” of Catanzaro. All subjects met the UK Parkinson's Disease Society Brain Bank clinical diagnostic criteria [7] for PD. DAT-SPECT imaging and cardiac MIBG scintigraphy were also performed in all patients to support a degenerative parkinsonian condition [8]. In all PD patients the RBD symptoms were assessed using the RBD Single-Question Screen [9] (RBD 1Q) a single “yes-no” question that queries the classic dream enactment behavior of RBD. We also performed in all PD patients (with and without RBD) an audio-visual polysomnography (PSG) to confirm the clinical diagnosis of RBD symptoms in PD-RBD group and exclude their presence in PD group. On PSG, a prominent muscle activity in REM sleep associated with abnormal behaviors was required to confirm the clinical diagnosis of RBD [10]. Exclusion criteria were: 1) current treatment with medications know modify REM sleep architecture and muscle tone as serotonin reuptake inhibitors; 2) evidence of structural abnormalities in the brain affecting the gray matter; 3)head movement artifacts during the MRI section; 4) evidence of global cognitive impairment (MMSE 26) [11]. Before MRI examination, patients underwent a careful clinical assessment including Hoehn-Yahr (HeY) staging [12] Unified Parkinson's Disease Rating scale (UPDRS) scores [13] and neuropsychological evaluation including MMSE, verbal fluency, token test, Frontal Assessment Battery (FAB) and Back Depression Inventory. Patients were compared to eighteen age-and sexmatched control subjects who had no history suggestive of RBD or other neurological or psychiatric diseases and showed normal MRI of the brain, both normal DAT-SPECT and cardiac MIBG scintigraphy.
3. Magnetic resonance imaging Brain MRI was performed according to our routine protocol by a 3 T scanner with an 8-channel head coil (Discovery MR-750, GE, Milwaukee, WI, USA). Structural MRI data were acquired using a 3D T1-weighted spoiled gradient echo (SPGR) sequence with the following parameters: TR: 3.7 ms, TE: 9.2 ms, flip angle 12 , voxelsize 1 1 1 mm3. Subjects were positioned to lie comfortably in the scanner with a forehead-restraining strap and various foam pads to ensure head fixation. 4. Voxel-based morphometry Data were processed and examined using the SPM8 software (Wellcome Trust Centre for Neuroimaging, London, UK), where we applied VBM implemented in the VBM8 toolbox (http://dbm.neuro. uni-jena.de/vbm.html) with default parameters incorporating the DARTEL toolbox that was used to obtain a high-dimensional normalization protocol. Images were bias-corrected, tissue classified, and registered using linear (12-parameter affine) and non-linear transformations (warping), within a unified model [14]. Subsequently, the warped gray matter (GM) segments were affine transformed into Montreal Neurological Institute (MNI) space and were scaled by the Jacobian determinants of the deformations (modulated GM volumes). Finally, the modulated volumes were smoothed with a Gaussian kernel of 8 mm full width at half maximum (FWHM). The GM volume maps were statistically analyzed using the general linear model based on Gaussian random field theory. Statistical analysis
1005
consisted of an analysis of covariance (AnCOVA) used for investigating the main effect of group (F-test). The advantage of an SPM-related Fstatistic is that increases and decreases of GM volumes are analyzed together to detect morphological changes in three or more groups. Age and total intracranial volume (ICV) were included in the model as covariates of no-interest. Two approaches to statistical threshold maps were applied. First, we applied a conservative approach with a whole-brain statistical threshold correction (p < 0.05, family-wise error (FWE)). Second, since that in vivo evidence of GM abnormalities in PD patients with RBD has never been reported, the data were also presented by using a less-stringent, uncorrected threshold (p < 0.001, cluster (k) threshold ¼ 10 voxels) to detect subtle morphological changes. 4.1. Automated subcortical volumetry To corroborate voxel-based findings we further performed automated labeling and quantification of subcortical brain regions volume using FreeSurfer 5.0. The automated procedures for volumetric measures of several deep gray matter structures have been previously described [6,15]. The automated subcortical segmentation performed by FreeSurfer required these steps: Firstly, an optimal linear transform is computed that maximizes the likelihood of the input image, given an atlas constructed from manually labeled images. A non-linear transform is then initialized with the linear one, and the image is allowed to further deform to better match the atlas. Finally, a Bayesian segmentation procedure is performed, and the maximum a posteriori estimate of the labeling is computed. This approach provides advantages similar to manual ROI drawing [6,16] without the potential for rater bias, offering an anatomically accurate rendering of regional volumes. ICV was Table 1 Demographic and clinical characteristics and neuroimaging features of PD patients and controls. Variables
PD-RBD (n ¼ 11)
Age (y) Sex (m/w) Age at onset (y) Disease duration (y) Hoehn and Yahr (HY) stage UPDRS UPDRS-ME Levodopa dosage (mg/die) Rigid-akinetic form (n) Onset of RBD symptoms (y) Cognitive data MMSE Verbal fluency Token test F.A.B. Beck depression inventory DAT-SPECT Put/Cau right Put/Cau left Cardiac MIBG scintigraphy H/M ratio of delayed images
66.6 8/3 61.9 4.72 1.95
± 7.4 ± 6.4 ± 4.07 ± 0.57
PD (n ¼ 11) 66.9 8/3 62.5 4.36 1.86
± 7.9 ± 7.27 ± 4.2 ± 0.59
Controls (n ¼ 18)
P-value
65.1 ± 7.8 13/5 e e e
0.79a
0.85b 0.70b 0.92c
30 ± 15.4 21.6 ± 10.4 563.6 ± 167.3
28.9 ± 10.9 19.9 ± 10.3 547.7 ± 155.4
e e e
3/11
2/11
e
0.83b 0.69c 0.64c
2.95 ± 1.09
27.4 24.4 30.4 13.6 7.45
± ± ± ± ±
2.87 8.77 5.03 1.98 4.01
26.9 24.6 29.5 14.3 7.54
± ± ± ± ±
3.27 12.0 2.08 2.36 3.72
27.9 27.2 31.0 15.2 6.50
± ± ± ± ±
1.85 7.38 1.88 1.74 6.89
0.86d 0.65a 0.14d 0.09a 0.31d
1.70 ± 0.22 1.73 ± 0.19
1.66 ± 0.21 1.64 ± 0.19
2.46 ± 0.09 2.45 ± 0.11
