- Email: [email protected]
Transcallosal diffusion tensor abnormalities in predominant gait disorder parkinsonism
Parkinsonism and Related Disorders 20 (2014) 53e59
Contents lists available at ScienceDirect
Parkinsonism and Related Disorders journal homepage: www.elsevier.com/locate/parkreldis
Transcallosal diffusion tensor abnormalities in predominant gait disorder parkinsonism Ling-Ling Chan a, Kia-Min Ng a, e, Helmut Rumpel a, Stephanie Fook-Chong c, Hui-Hua Li c, Eng-King Tan b, c, d, e, * a
Department of Diagnostic Radiology, Singapore General Hospital, Republic of Singapore Department of Neurology, Singapore General Hospital, Republic of Singapore Department of Clinical Research, Singapore General Hospital, Republic of Singapore d National Neuroscience Institute, Republic of Singapore e Duke-NUS Graduate Medical School, Republic of Singapore b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 July 2013 Received in revised form 12 September 2013 Accepted 13 September 2013
Background: There have been no previous diffusion tensor imaging (DTI) studies comparing Parkinson’s disease (PD) with postural instability and gait disorder (PIGD) parkinsonism. Objective: Utilizing DTI with 2-region tractography, we conducted a case control study to determine if different brain regions representing the neural network of the motor system are differentially affected in PIGD compared to PD and controls. Methods: On a 3 T MR machine, using manual ROI (regions of interest) we determined the fractional anisotropy (FA) and apparent diffusion coefficient (ADC) values on DTI in anatomical brain regions representing the extrapyramidal, pyramidal, and transcallosal tracts, aided by 2-region tractography. FA and ADC were correlated with the Tinetti score (measure of gait and balance). Results: Sixty-five subjects (21 PD, 25 PIGD, 19 controls) were included in the analysis. We demonstrated greater ADC abnormalities in the extrapyramidal, pyramidal and transcallosal motor systems in PIGD compared to controls. Multivariate analysis taking into consideration various clinical variables showed that the FA (p ¼ 0.02) and ADC (p ¼ 0.001) values in the corpus callosum body differentiated PIGD from PD. PIGD with low Tinetti score had a lower FA (p ¼ 0.02) and a higher ADC value (corpus callosum body) (p ¼ 0.03) compared to those with a high score. Conclusions: We demonstrated for the first time that DTI abnormalities along the transcallosal motor tract in the body of the corpus callosum, but not the substantia nigra, differentiated PIGD from PD, and the degree of corpus callosum body abnormality correlated with the Tinetti score (a measure of risk of falls). Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Parkinson’s disease Diffusion tensor imaging Postural instability Gait disorder
1. Introduction Gait disorders are common in the elderly population and cerebral small vessel disease has been shown to be related to gait disturbances [1]. Patients presenting with predominant gait disorder such as postural instability, frequent falling and freezing are frequently encountered in Neurology Outpatient Clinics. Patients with postural instability and gait disorder (PIGD), thought to be a subtype of Parkinson’s disease (PD) present with predominant gait disorder. In a clinico-pathological study, PIGD patients tend to be * Corresponding author. Department of Neurology, Singapore General Hospital, Outram Road, Singapore 169608, Republic of Singapore. E-mail address: [email protected] (E.-K. Tan). 1353-8020/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.parkreldis.2013.09.017
less levodopa responsive [2] and have a higher association with cognitive dysfunction and leukoaraiosis on brain imaging [3,4]. It is well established that symptoms of parkinsonism are a consequence of decreased striatal dopamine levels arising from selective and progressive loss of dopaminergic cells within the pars compacta of the substantia nigra and locus ceruleus of the mid-brain. Levodopa replacement or deep brain stimulation is effective in alleviating these symptoms [5,6]. However, symptoms related to PIGD are generally less amenable to medical and surgical therapy, and selection of optimal surgical targets for gait abnormality has attracted considerable research interest. Various hypotheses explaining these discrepant observations have suggested basal ganglia infarcts, deep white matter ischemia and even corpus callosal atrophy.
54
L.-L. Chan et al. / Parkinsonism and Related Disorders 20 (2014) 53e59
Fig. 1. DTI fiber tracking using 2-region tractography on FA color maps. Anatomic sulcal morphology (A) aided placement of 2 superior seed ROIs over the motor strip (B). Inferior seed ROIs were drawn in the blue regions of the ventral pons (C, D). Fiber tracts connecting ipsilateral seed ROIs e the corticospinal tract, CST (thick arrow), and both superior ROIs e transcallosal motor tracts, CCm (thin arrow), are depicted in 3D (E) and coronal 2D view (inset). (CS: central sulcus, pCS: pre-central sulcus, SFS: superior frontal sulcus). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Diffusion tensor imaging (DTI), a non-invasive MR imaging modality, has been shown to be potentially useful in differentiating PD from healthy controls [7e9]. In neural fiber tracts, water diffuses asymmetrically, being fastest along the major axis parallel to the direction of the fibers. The DTI indices of apparent diffusion coefficient (ADC) and fractional anisotropy (FA) describe the magnitude (diffusivity) and directionality (anisotropy) of diffusion of water molecules in the brain respectively [10,11]. Increased ADC and reduced FA values reveal underlying histopathological processes such as gliosis, demyelination, edema and axonal loss. Hence, they are used to elucidate the integrity of tissue microstructure in the brain. Fiber tracking, or tractography, uses the similarity in directionality of anisotropic diffusion between adjacent voxels on DTI to demonstrate neuronal projections from a single or connecting two functional center(s). Using two-region DTI tractography, the corticospinal and transcallosal fibers of the motor pathway can be localized and differentiated in the subcortical white matter. The objective of this case control DTI study, aided by two region tractography, is to determine if different anatomical brain regions representing the extrapyramidal and pyramidal motor systems,
transcallosal motor connection and corpus callosum are differentially affected in patients with PIGD parkinsonism compared to typical PD, and controls. 2. Methods 2.1. Study subjects All study subjects gave written informed consent. Approval from the institutional ethics committee was obtained for this study. Patients who presented initially with rest tremor, bradykinesia and rigidity and diagnosed with PD based on the United Kingdom PD Brain Bank clinical criteria [12], by a movement disorders neurologist in a tertiary referral center were included. Parkinsonian patients who presented initially with predominant postural instability, frequent falls, freezing and walking difficulty were defined as PIGD parkinsonism following criteria used in the literature [13]. Exclusion criteria included those with evidence of cognitive dysfunction (based on mini mental state score), severe joint problems or structural lower limb abnormalities, hemiplegia or focal neurological deficits to suggest a stroke, or neurological signs suggestive of normal pressure hydrocephalus, progressive supranuclear gaze palsy or multiple system atrophy. For every PD patient, we sought a PIGD patient and control of similar age and gender. The controls were healthy volunteers with no evidence of PD, recruited over the same period. A history of common vascular risk factors, such as hypertension, diabetes or hyperlipidemia, was not an exclusion criterion for all the study subjects.
L.-L. Chan et al. / Parkinsonism and Related Disorders 20 (2014) 53e59
55
Fig. 2. FA color maps depicting placement of circle ROIs representing the extrapyramidal, pyramidal corticospinal (CST) and corpus callosal (CC) tracts in the (A) pons CST, (B) substantia nigra and cerebral peduncle CST, (C) putamen, thalamus and internal capsule CST, (D) corpus callosum genu and splenium, (E) corona radiata CST and corpus callosum body (CCm) and (F) centrum semiovale CST and CCm. Identification of the CST and CC motor tracts at the corona radiata (E) and centrum semiovale (F) for ROI placement was aided by the tract volumes embedded on the b0 diffusion images (images on right). To evaluate the risk of falling, all study subjects were assessed using the Tinetti gait and balance test [14]. This comprised of a balance (9 items) and a gait (7 items) section. Each item was scored using a scale of 0, 1 or 2 and the total possible score was 28, with a lower score indicating more severe gait impairment. In addition, information on age, gender, age of onset of disease, family history, and severity of PD (Hoehn and Yahr scale, H&Y) were collected. A total of 100 study subjects were screened, of which 35 were excluded because they did not satisfy the inclusion and exclusion criteria or were unwilling to participate in study/unable to complete the MRI examination. 2.2. MR imaging The MR scans were performed on a 3 T Scanner (Siemens Trio, Erlangen, Germany), using a 12 channel phased array head coil. To reduce head motion, the subjects’ head was stabilized using Velcro straps, foam supports and padding, and patients were scanned during their “on” state. The scans were acquired in the axial plane, but tilted to the anterior commissureeposterior commissure (ACePC) line. The DTI scan was a spin-echo echo planar imaging (SE-EPI) sequence, using the following parameters: repetition time, TR ¼ 8200 ms; echo time, TE 86 ms; two b values (0 and 800 s/mm2); diffusion sensitization in 30 non-collinear directions; inplane resolution of 1.9 1.9 mm (FOV 240 240 mm, matrix 128 128); 64 contiguous 2 mm slices; and iPAT factor 2. The other pulse sequences were: (1) T2weighted (TR/TE 4530/84 ms, matrix 448 448), (2) MPRAGE with (TR/TE/TI 2200/ 3.0/900 ms), flip angle of 9 , matrix 256 256, and [3] fluid attenuated inversion recovery (FLAIR) with TR/TE/TI 7150/74/2535 ms and matrix 256 256. 2.3. DTI fiber tracking Three-dimensional (3D) reconstructions of fiber tracts were performed using the commercially available syngo DTI Tractography software (Siemens Healthcare Sector, Erlangen, Germany). The tractography algorithm is based on a deterministic streamline (FACT) approach and tracts are calculated with a Runge Kutta 4th order optimization. Tracking was terminated when the FA value dropped below 0.2 or if the curvature exceeded 30 between two contiguous eigen-vectors, and the step length was 1 mm [15,16]. Two seed regions of interest (ROIs) were drawn in each cerebral hemisphere on the 2-dimensional (2D) fractional anisotropy color maps (Fig. 1). The superior seed ROI was manually placed over the motor strip, based on anatomic sulcal morphology at the vertex and in correlation with the motor activation areas on fMRI. The inferior seed area was drawn in the blue portions of the ventral pons. Fiber tracts passing through ipsilateral superior and inferior seed ROIs descended through the known corticospinal tract (CST). Fiber tracts connecting the two superior seed ROIs in the motor strips traverse the body of the corpus callosum and were designated transcallosal motor tracts (CCm). 2.4. Image analysis FA maps with directions encoded in color (red (in web version): lefteright, blue (in web version): superioreinferior, green (in web version): anterioreposterior),
trace-ADC maps, trace-diffusion weighted images, and b0 diffusion images were displayed side by side. For the FA and ADC analysis, circle ROIs (Fig. 2) were drawn over various brain regions on the FA maps and ADC maps based on anatomical knowledge and also aided by superimposing the 3D CST and CCm tract volumes on structural b0 diffusion images (Fig. 2E,F). These ROIs yielded averaged values of FA and ADC. Anatomical brain regions (Fig. 2) representing the (i) extrapyramidal system: substantia nigra, putamen, and ventrolateral thalamus, (ii) pyramidal system comprising corticospinal tract (CST) at these levels: pons, cerebral peduncle, posterior limb of the internal capsule, corona radiata and centrum semiovale, and (iii) corpus callosal tracts (CC) at the genu, splenium, and transcallosal motor connection (CCm) at the corpus callosum body and centrum semiovale were studied using ROIs volumes of 52.8 mm3 for the pons, putamen and thalamus, and 22.1 mm3 for all other structures. ROIs were drawn within the core of each structure on representative sections at 32 mm (pons), 12 mm (substantia nigra and cerebral peduncles), þ6 mm (putamen, thalamus, posterior limb of internal capsule), þ12 mm (genu and splenium of corpus callosum), þ26 mm (corona radiata and corpus callosum body) and þ36 mm (centrum semiovale) from the section containing the ACePC. Presence of infarcts on MRI were assessed by our neuroradiologist (LL-C) based on conventional T1- and T2-weighted MR sequences. ROIs were not taken from any areas of infarcts and did not affect the DTI measurements.
3. Statistical analysis Statistical analyses were carried out using the SPSS software. Pearson’s chi square test was used to compare the categorical variables and Student’s t-test for the continuous variables. The mean of the right and left FA and ADC values of various anatomical regions was used in the analysis. A logistic regression to distinguish between PD and PIGD was carried out with the following independent variables: FA/ADC value for the various brain regions, patients’ demographics, Tinetti score, and MRI infarcts. These were investigated for entry in the model by stepwise variable selection method. The variable entry and exit criteria were p < 0.05 and p > 0.10 respectively. Box plot was performed to display the distribution of corpus callosum (body) FA and ADC values among patients with different Tinetti score. Receiver operating characteristics (ROC) analysis was carried out to evaluate the sensitivity and specificity of the DTI parameters as a screening and diagnostic tool for PD. The area under the curve (AUC) highlighted the discriminative property of the scale. All statistical tests were evaluated at two-sided and statistical significance was at p < 0.05.
56
L.-L. Chan et al. / Parkinsonism and Related Disorders 20 (2014) 53e59
4. Results
5. Discussion
A total of 65 subjects including 25 PIGD, 21 PD and 19 controls participated. All three groups were matched for age and gender and their demographics were summarized in Table 1. The mean Tinetti score in PIGD and PD was significantly lower than controls, and PIGD had the lowest score among the 3 groups (Table 1). For common vascular risk factors, PIGD had the lowest frequency of hypertension but the highest frequency for diabetes mellitus and heart disease and MRI evidence of vascular infarcts (Table 1). The mean FA and ADC values of each of the brain region in the 3 groups of subjects were tabulated in Tables 2 and 3 respectively. The FA of the substantia nigra was lower in PD and PIGD compared to controls but the difference was not significant. However, the ADC value was higher in PD vs control (p ¼ 0.004) and PIGD vs control (p ¼ 0.046), but difference between PD vs PIGD was not significant. The ADC values were also greater in the thalamus (p ¼ 0.002, p ¼ 0.003), putamen (p ¼ 0.176, p ¼ 0.03) and CST at the centrum semiovale (p ¼ 0.038, p ¼ 0.479) and corona radiata (p ¼ 0.027, p ¼ 0.072) in PIGD vs controls and in PIGD vs PD. The FA value of the corpus callosum body (p ¼ 0.047) and corpus callosum genu (p ¼ 0.013) was lower in PIGD vs controls (p ¼ 0.047), and lower in PIGD vs PD for the corpus callosum body (p ¼ 0.008). The ADC values were greater in the corpus callosum splenium (p ¼ 0.087, p ¼ 0.018), body (p ¼ 0.004, p ¼ 0.001) and CCm in the centrum semiovale (p ¼ 0.003, p ¼ 0.01) in PIGD vs controls and in PIGD vs PD. In the multivariate analysis, after adjustments were made for the effects of the FA and ADC values for the various brain regions, vascular infarcts on MRI, and Tinetti score, the FA (p ¼ 0.023) and ADC (p ¼ 0.001) values in the corpus callosum body (CCm), but not in the other brain regions, differentiated between PIGD and PD. For PIGD patients, the FA value was lower for those with Tinetti score less than or equal to 16 compared to those more than 16 (p ¼ 0.02) (Fig. 3). The ADC value was higher for those with Tinetti score less than or equal to 16 compared to those more than 16 (p ¼ 0.03) (Fig. 4). For ADC at the corpus callosum body, the AUC was at 82.3%, and at cut-off of 760 106 mm2/s, the sensitivity was 76% and specificity was 71.4% to differentiate PIGD from PD. For FA, the AUC was at modest at 70.3%, and at cut-off of 870 103, the sensitivity was 68% and specificity was 62.0%. There was no specific FA or ADC value which gave a 100% sensitivity and specificity.
While studies have suggested gait symptoms in PD are less levodopa responsive and amenable to DBS surgery [5,6] the exact pathophysiologic differences between PD and PIGD are still unclear. A recent MR DTI in vascular parkinsonism revealed frontal lobe white matter microstructural disruption [17]. In our case control DTI tractrographic study comparing PIGD patients with PD and controls, we demonstrated DTI abnormalities in the substantia nigra, thalamus, CST, corpus callosum tracts and CCm in PIGD compared to controls. The abnormality in the substantia nigra was compatible with the levodopa responsiveness of these patients. Though the mean ADC values were higher in the thalamus, putamen, corpus callosum and CCm in PIGD compared to PD, multivariate analysis revealed that only the FA and ADC values in the corpus callosum body differentiated PIGD from PD. A previous DTI-based tractography study in healthy individuals identified numerous vertical segments of the corpus callosum, containing fibers projecting into prefrontal, premotor (and supplementary motor), primary motor and sensory areas, parietal, temporal, and occipital cortical areas [18]. Transcallosal motor fiber bundles were found to cross the corpus callosum in posterior part of body and splenium. The anatomic projections delineated in this study are compatible with our study observations that corpus callosum body abnormalities differentiate patients with predominant postural instability and gait disorder compared to PD and healthy controls. Cortical projections to the premotor and supplemental motor areas, and primary motor and sensory areas arise from the body of the corpus callosum. Clinical, neurophysiologic and imaging studies have suggested that disruptions of the cortico-basal ganglia and basal gangliaebrainstem networks and dysfunction of sensorimotor integration processes can lead to gait disturbance [19]. Our study provides evidence that differential disruption of the callosal projections to these motor and sensory areas rather than primary abnormalities in the basal ganglia could explain why some parkinsonism patients developed postural instability and gait problems more than others. In the DATATOP cohort, PIGD symptoms were more common at onset in PD patients with a rapid rate of disease progression and reported greater subjective intellectual and motor impairment than the tremor onset PD [13]. The greater impairment in the corpus callosum in PIGD as highlighted in our study suggests either that the degree of disease progression is greater or mixed pathologies are present compared to the PD.
Table 1 Clinical features of PD, PIGD and controls. PD
Number Gender Age (years) Tinetti score Balance Gait Total H&Y Vascular risk factors Diabetes Hypertension Heart disease Cholesterol MRI vascular infarcts *p < 0.05.
PIGD
Controls
p Value PD vs controls
PIGD vs controls
PD vs PIGD
e e
e e
21 17 males 4 females 72.0 4.8 (62.0, 82.0)
25 18 males 7 females 73.3 6.2 (63.0, 87.0)
19 16 males 3 females 71.5 5.0 (62.0, 80.0)
e e 0.967
0.543
11.2 2.9 (4, 15) 8.6 1.9 (3, 11) 19.9 4.4 (7, 24) 2.2 0.5 (2, 4)
9.3 3.4 (3, 16) 6.8 2.5 (2, 12) 16.1 5.3 (7, 26) 3.0 1.0 (2, 4)
15.3 0.7 (13, 16) 11.4 0.8 (10, 12) 26.7 1.2 (24, 28) e
<0.01* <0.01* <0.01*
<0.01* <0.01* <0.01*
5 (23.8%) 12 (57.1%) 3 (14.3%) 9 (42.9%) 4 (19.0%)
10 (40%) 13 (52%) 7 (28%) 11 (44%) 15 (60%)
3 (15.8%) 12 (63.2%) 1 (5.3%) 6 (31.6%) 5 (26.3%)
0.527 0.962 0.342 0.462 0.587
0.081 0.697 0.053 0.402 0.028*
0.690 0.047* 0.010* 0.010*
0.243 0.727 0.261 0.938 0.005*
L.-L. Chan et al. / Parkinsonism and Related Disorders 20 (2014) 53e59
57
Table 2 FA values of study subjects. ROI (103)
PD mean SD (range)
Extrapyramidal system Substantia nigra 519.1 101.1 (328.4, Thalamus 328.5 42.9 (239.6, Putamen 268.2 63.1 (122.6, Pyramidal system e corticospinal tract (CST) Pons 578.5 42.7 (504.5, Cerebral peduncle 714.7 52.3 (620.7, Internal capsule 763.7 36.9 (712.7, Corona radiata 558.2 59.5 (426.3, Centrum semiovale 413.2 76.3 (256.3, Corpus callosum tracts Genu 859.6 40.3 (757.0, Splenium 918.7 33.9 (849.3, Body (CCm) 875.9 60.4 (707.7, Centrum semiovale 480.0 84.3 (368.3, (CCm)
PIGD mean SD (range)
752.0) 384.9) 371.0)
521.8 80.6 (375.2, 667.5) 306.4 45.5 (209.6, 393.6) 260.14 67.5 (157.7, 384.0)
669.8) 839.8) 847.9) 694.0) 558.9) 930.7) 984.5) 949) 629.4)
Controls mean SD (range)
575.7 714.3 768.0 576.0 433.5
76.0 50.8 38.8 69.8 86.5
818.8 102.4 865.8 120.1 814.3 71.2 462.9 117.3
(428.2, (561.3, (698.1, (441.1, (260.9,
774.0) 805.8) 838.5) 735.4) 645)
(477.5, 913.3) (393, 977) (674, 940.5) (214.1, 795.2)
The Tinetti scale is a sensitive, reliable and valid clinical instrument for assessing the risk of falling and is a good measure of balance and gait in the elderly. It has also been shown to correlate with gait and postural stability UPDRS scores in PD patients [14]. PIGD patients are at a very high risk for falls and thus their Tinetti scores will be particularly relevant as a measure of the clinical severity. In our study, there was a lower FA value and a higher ADC value (corpus callosum body) in PIGD patients with low Tinetti score compared to those with a high score (Figs. 3 and 4). This provides further supporting evidence that the corpus callosum abnormality is associated with gait dysfunction. While the correlation of a reduced FA value or increased ADC value with degeneration of specific pathological structures (active neuronal loss, gliosis or inflammation) is not entirely clear, these changes in the corpus callosum probably represent a combination of microstructural damage and cortical atrophy with secondary transcallosal degeneration. In PD, the substantia nigra is the most severely affected primary site of pathology and secondary degeneration could be present in the basal ganglia and extra-striatal regions in the later stages. In a clinico-pathologic study, the majority of tremor onset PD cases had Lewy body pathology while the majority of PIGD-onset cases had other variable forms of pathology [2]. The differences in the ADC and FA values in the substantia nigra were greater in PD vs controls
PD vs controls
PIGD vs controls
PD vs PIGD
0.780 0.784 0.978
0.824 0.678 0.812
0.993 0.262 0.908
634.7) 765.2) 819.8) 747.4) 531.8)
0.891 0.531 0.393 0.922 0.966
0.948 0.523 0.203 0.878 0.818
0.984 0.999 0.927 0.636 0.648
(806.3, 976) (836.3, 963) (771, 959.3) (320.8, 822.8)
0.567 0.878 0.123 0.998
0.013* 0.218 0.047* 0.825
0.114 0.07 0.008* 0.849
536.7 58.3 (436.5, 678.2) 318.5 53.2 (239.4, 456.9) 272.3 62.9 (131.1, 411.0) 570.5 698.6 747.4 566.2 419.3
27.5 34.3 41.6 67.9 64.4
882.8 45.3 906.5 33.1 857.0 46.7 482.0 111.1
p Value
(523.8, (651.0, (690.9, (464.4, (314.7,
compared to PIGD vs controls, and the absolute values of FA and ADC were lower and higher in PD compared to PIGD, suggesting more severe involvement of the substantia nigra in PD compared to PIGD. Our study showed that vascular infarcts were more commonly seen in PIGD compared to PD and controls. It is possible that vascular lesions contribute to the pathophysiology and the clinical phenotype of patients with PIGD may overlap with “vascular” form of parkinsonism. Specific neuronal structures or tracks could have been dysregulated as a consequence of vascular damage. Recent structural MR studies have also showed a higher prevalence of leukoaraiosis in PIGD compared to PD. A previous structural MRI study and meta-analysis of published literature has examined corpus callosum as a marker of cortical pathology in neurodegenerative diseases [20] and concluded that PD patients do not have callosal atrophy compared to other neurodegenerative diseases. However, it unclear if any of these patients belonged to the PIGD phenotype. In a recent study involving a general elderly population, DTI was used to examine the genu and splenium of the corpus callosum only, an independent association of gait with genu abnormalities and a correlation of the DTI abnormality with the validated Tinetti gait scores was found [1]. DTI abnormalities in the corpus callosum have been shown to be present in corticobasal syndrome [21,22] and the anterior corpus callosum is more affected
Table 3 ADC values of study subjects. ROI (106 mm2/s)
PD mean SD (range)
Extrapyramidal system Substantia nigra 901.6 148.4 Thalamus 749.1 67.9 Putamen 793.4 129.7 Pyramidal system e corticospinal tract (CST) Pons 798.8 57.8 Cerebral peduncle 942.5 123.4 Internal capsule 732.2 38.9 Corona radiata 755.6 40.4 Centrum semiovale 816.0 66.6 Corpus callosum tracts Genu 759.2 72.6 Splenium 671.4 86.7 Body (CCm) 717.8 77.9 Centrum semiovale (CCm) 847.7 82.5 *p < 0.05.
(564.7, 1164.5) (649.8, 928.7) (651.4, 1207.8) (677.9, (666.5, (643.1, (685.8, (679.9,
909.1) 1173.8) 788.4) 828.5) 942.7)
(553.0, (504.8, (586.7, (715.9,
905.0) 856) 894.7) 1076.7)
PIGD mean SD (range)
856.1 116.8 (569.5, 1039.3) 841.3 122.2 (678.8, 1140.3) 986.9 353.7 (480.4, 2239) 785.6 62.2 1000.3 164.9 741.3 59.1 802.9 88.7 844.9 98.1 859.2 782.0 876.9 961.3
280.0 179.6 134.6 178.7
(660.6, (762.1, (631.3, (636.4, (692.2,
884.7) 1588.4) 864.5) 981.4) 1063.0)
(99.8, 1727.5) (565, 1437.7) (665, 1192) (648.0, 1326.5)
Controls mean SD (range)
p Value PD vs controls
PIGD vs controls
PD vs PIGD
746.0 182.3 (105.8, 1005.5) 741.4 57.8 (636.4, 866.0) 849.0 176.8 (541.9, 1267.8)
0.004* 0.961 0.765
0.046* 0.002* 0.176
0.558 0.003* 0.030*
773.8 43.4 918.2 203.8 728.2 36.0 745.4 72.2 780.6 80.8
0.341 0.889 0.960 0.893 0.383
0.769 0.242 0.630 0.027* 0.038*
0.705 0.472 0.791 0.072 0.479
0.876 0.840 0.096 0.860
0.066 0.087 0.004* 0.003*
0.171 0.018* 0.001* 0.010*
730.4 695.1 766.6 826.5
89.8 95.4 88.5 77.0
(669.6, (200.1, (672.2, (653.3, (692.3,
841.7) 1176.5) 815.8) 956.9) 1021.9)
(547.3, 874.3) (546.7, 857.7) (660, 993.7) (715.7, 954.7)
58
L.-L. Chan et al. / Parkinsonism and Related Disorders 20 (2014) 53e59
clinical trials or for prognostic purposes. Second; there is no gold standard validation of the diagnosis of PIGD and it is possible that this entity is clinically heterogeneous even though gait abnormality is the predominant manifestation. Hence there may be overlap with “vascular” parkinsonism or even some of the atypical parkinsonian disorders. Third; as there is paucity of white matter tracts in the deep gray matter structures, the FA values in these structures in our study have to be interpreted with caution. In conclusion, we demonstrated for the first time that DTI abnormalities along the transcallosal motor tract in the body of the corpus callosum, but not the substantia nigra, differentiated PIGD from PD, and the degree of corpus callosum body abnormality correlated with the Tinetti score (a measure of risk of falls). Author contributions
Fig. 3. For PIGD patients, the FA value was lower for those with Tinetti score less than or equal to 16 compared to those more than 16 (p ¼ 0.02).
in progressive supranuclear palsy compared to PD [23]. These associations bear some similarities to PIGD. As PIGD may have varied pathologies and clinical associations, it is conceivable that the pathophysiologic process may overlap with some of the atypical parkinsonian syndromes. To assess the potential utility of the FA and ADC values in clinical practice, we conducted a ROC analysis to assess the sensitivity and specificity of these values. Ideal screening and diagnostic tests should have both high positive and negative predictive values. As expected, there was a considerable overlap of the FA and ADC values (corpus callosum) between PD and PIGD. This is to be expected because of the considerable overlap in gait dysfunction. For example, the area under curve for the ADC (corpus callosum body) was 82.3%, and with a cut-off value of 760 106 mm2/s the sensitivity was 76% and specificity 71.4%. While these values cannot be used for screening, their clinical utility should be addressed in a long term longitudinal study. Our study has limitations. First; the cross-section design did not allow measurement of FA or ADC values serially for individual subjects in a longitudinal manner. A prospective study to measure the DTI values in the corpus callosum would be useful to monitor the rate of decline in these patients. This might provide corroborative data when monitoring patients’ response to neuroprotective
Fig. 4. For PIGD patients, the ADC value was higher for those with Tinetti score less than or equal to 16 compared to those more than 16 (p ¼ 0.03).
EK Tan, LL Chan designed the study. KM Ng, H Rumpel, EK Tan, LL Chan obtained the data. HH Li, S Fook-Chong, KM Ng performed the statistical analysis. LL Chan and EK Tan wrote the manuscript. All the authors were involved in the interpretation of the data and revision of the manuscript. LL Chan supervised the study. Financial disclosures All the authors have no conflict of interest and financial disclosures related to this study to declare. The corresponding author declares that the Authors take full responsibility for the data, the analyses and interpretation, and the conduct of the research; that the Author has full access to all of the data; and that the Author has the right to publish any and all data, separate and apart from the guidance of any sponsor. This work is supported by the National Medical Research Council, Duke-NUS Graduate Medical School, Singapore Millennium Foundation, SingHealth Foundation. Acknowledgments We thank the National Medical Research Council, Singapore Millennium Foundation, SingHealth Foundation, and Duke NUS Graduate Medical School for their support. References [1] Bhadelia RA, Price LL, Tedesco KL, Scott T, Qiu WQ, Patz S, et al. Diffusion tensor imaging, white matter lesions, the corpus callosum, and gait in the elderly. Stroke 2009;40:3816e20. [2] Rajput AH, Pahwa R, Pahwa P, Rajput A. Prognostic significance of the onset mode in parkinsonism. Neurology 1993;43:829e30. [3] Lee SJ, Kim JS, Lee KS, An JY, Kim W, Kim YI, et al. The severity of leukoaraiosis correlates with the clinical phenotype of Parkinson’s disease. Arch Gerontol Geriatr 2009;49:255e9. [4] Alves G, Larsen JP, Emre M, Wentzel-Larsen T, Aarsland D. Changes in motor subtype and risk for incident dementia in Parkinson’s disease. Mov Disord 2006;21:1123e30. [5] Jankovic J, Aguilar LG. Current approaches to the treatment of Parkinson’s disease. Neuropsychiatr Dis Treat 2008;4:743e57. [6] Lang AE, Lozano AM. Parkinson’s disease. First of two parts. N Engl J Med 1998;339:1044e53. [7] Vaillancourt DE, Spraker MB, Prodoehl J, Vaillancourt DE, Spraker MB, Prodoehl J, et al. High-resolution diffusion tensor imaging in the substantia nigra of de novo Parkinson disease. Neurology 2009;72:1378e84. [8] Gattellaro G, Minati L, Grisoli M, Mariani C, Carella F, Osio M, et al. White matter involvement in idiopathic Parkinson disease: a diffusion tensor imaging study. AJNR Am J Neuroradiol 2009;30:1222e6. [9] Chan LL, Rumpel H, Yap K, Lee E, Loo HV, Ho GL, et al. Case control study of diffusion tensor imaging in Parkinson’s disease. J Neurol Neurosurg Psychiatry 2007;78:1383e6. [10] Basser PJ, Pierpaoli C. Microstructural and physiological features of tissues elucidated by quantitative-diffusion-tensor MRI. J Magn Reson B 1996;111: 209e19.
L.-L. Chan et al. / Parkinsonism and Related Disorders 20 (2014) 53e59 [11] Beaulieu C. The basis of anisotropic water diffusion in the nervous system e a technical review. NMR Biomed 2002;15:435e55. [12] 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):181e4. [13] Jankovic J, McDermott M, Carter J, Gauthier S, Goetz C, Golbe L, et al. Variable expression of Parkinson’s disease: a base-line analysis of the DATATOP cohort. The Parkinson Study Group. Neurology 1990;40:1529e34. [14] Kegelmeyer DA, Kloos AD, Thomas KM, Kostyk SK. Reliability and validity of the Tinetti mobility test for individuals with Parkinson disease. Phys Ther 2007;87:1369e78. [15] Wakana S, Jiang H, Nagae-Poetscher LM, van Zijl PC, Mori S. Fiber tract based atlas of human white matter anatomy. Radiology 2004;230:77e87. [16] Nucifora PG, Verma R, Lee SK, Melhem ER. Diffusion-tensor MR imaging and tractography: exploring brain microstructure and connectivity. Radiology 2007;245:367e84. [17] Wang HC, Hsu JL, Leemans A. Diffusion tensor imaging of vascular parkinsonism: structural changes in cerebral white matter and the association with clinical severity. Arch Neurol 2012:1e9 [doi].
59
[18] Hofer S, Frahm J. Topography of the human corpus callosum revisitede comprehensive fiber tractography using diffusion tensor magnetic resonance imaging. Neuroimage 2006;32:989e94. [19] Takakusaki K, Tomita N, Yano M. Substrates for normal gait and pathophysiology of gait disturbances with respect to the basal ganglia dysfunction. J Neurol 2008;255:19e29. [20] Wiltshire K, Foster S, Kaye JA, Small BJ, Camicioli R. Corpus callosum in neurodegenerative diseases: findings in Parkinson’s disease. Dement Geriatr Cogn Disord 2005;20:345e51. [21] Boelmans K, Bodammer NC, Suchorska B, Kaufmann J, Ebersbach G, Heinze HJ, et al. Diffusion tensor imaging of the corpus callosum differentiates corticobasal syndrome from Parkinson’s disease. Parkinsonism Relat Disord 2010;16: 498e502. [22] Boelmans K, Kaufmann J, Bodammer N, Ebersbach G, Behlau G, Heinze HJ, et al. Involvement of motor pathways in corticobasal syndrome detected by diffusion tensor tractography. Mov Disord 2009;24:168e75. [23] Ito S, Makino T, Shirai W, Hattori T. Diffusion tensor analysis of corpus callosum in progressive supranuclear palsy. Neuroradiology 2008;50:981e5.
Contents lists available at ScienceDirect
Parkinsonism and Related Disorders journal homepage: www.elsevier.com/locate/parkreldis
Transcallosal diffusion tensor abnormalities in predominant gait disorder parkinsonism Ling-Ling Chan a, Kia-Min Ng a, e, Helmut Rumpel a, Stephanie Fook-Chong c, Hui-Hua Li c, Eng-King Tan b, c, d, e, * a
Department of Diagnostic Radiology, Singapore General Hospital, Republic of Singapore Department of Neurology, Singapore General Hospital, Republic of Singapore Department of Clinical Research, Singapore General Hospital, Republic of Singapore d National Neuroscience Institute, Republic of Singapore e Duke-NUS Graduate Medical School, Republic of Singapore b c
a r t i c l e i n f o
a b s t r a c t
Article history: Received 23 July 2013 Received in revised form 12 September 2013 Accepted 13 September 2013
Background: There have been no previous diffusion tensor imaging (DTI) studies comparing Parkinson’s disease (PD) with postural instability and gait disorder (PIGD) parkinsonism. Objective: Utilizing DTI with 2-region tractography, we conducted a case control study to determine if different brain regions representing the neural network of the motor system are differentially affected in PIGD compared to PD and controls. Methods: On a 3 T MR machine, using manual ROI (regions of interest) we determined the fractional anisotropy (FA) and apparent diffusion coefficient (ADC) values on DTI in anatomical brain regions representing the extrapyramidal, pyramidal, and transcallosal tracts, aided by 2-region tractography. FA and ADC were correlated with the Tinetti score (measure of gait and balance). Results: Sixty-five subjects (21 PD, 25 PIGD, 19 controls) were included in the analysis. We demonstrated greater ADC abnormalities in the extrapyramidal, pyramidal and transcallosal motor systems in PIGD compared to controls. Multivariate analysis taking into consideration various clinical variables showed that the FA (p ¼ 0.02) and ADC (p ¼ 0.001) values in the corpus callosum body differentiated PIGD from PD. PIGD with low Tinetti score had a lower FA (p ¼ 0.02) and a higher ADC value (corpus callosum body) (p ¼ 0.03) compared to those with a high score. Conclusions: We demonstrated for the first time that DTI abnormalities along the transcallosal motor tract in the body of the corpus callosum, but not the substantia nigra, differentiated PIGD from PD, and the degree of corpus callosum body abnormality correlated with the Tinetti score (a measure of risk of falls). Ó 2013 Elsevier Ltd. All rights reserved.
Keywords: Parkinson’s disease Diffusion tensor imaging Postural instability Gait disorder
1. Introduction Gait disorders are common in the elderly population and cerebral small vessel disease has been shown to be related to gait disturbances [1]. Patients presenting with predominant gait disorder such as postural instability, frequent falling and freezing are frequently encountered in Neurology Outpatient Clinics. Patients with postural instability and gait disorder (PIGD), thought to be a subtype of Parkinson’s disease (PD) present with predominant gait disorder. In a clinico-pathological study, PIGD patients tend to be * Corresponding author. Department of Neurology, Singapore General Hospital, Outram Road, Singapore 169608, Republic of Singapore. E-mail address: [email protected] (E.-K. Tan). 1353-8020/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.parkreldis.2013.09.017
less levodopa responsive [2] and have a higher association with cognitive dysfunction and leukoaraiosis on brain imaging [3,4]. It is well established that symptoms of parkinsonism are a consequence of decreased striatal dopamine levels arising from selective and progressive loss of dopaminergic cells within the pars compacta of the substantia nigra and locus ceruleus of the mid-brain. Levodopa replacement or deep brain stimulation is effective in alleviating these symptoms [5,6]. However, symptoms related to PIGD are generally less amenable to medical and surgical therapy, and selection of optimal surgical targets for gait abnormality has attracted considerable research interest. Various hypotheses explaining these discrepant observations have suggested basal ganglia infarcts, deep white matter ischemia and even corpus callosal atrophy.
54
L.-L. Chan et al. / Parkinsonism and Related Disorders 20 (2014) 53e59
Fig. 1. DTI fiber tracking using 2-region tractography on FA color maps. Anatomic sulcal morphology (A) aided placement of 2 superior seed ROIs over the motor strip (B). Inferior seed ROIs were drawn in the blue regions of the ventral pons (C, D). Fiber tracts connecting ipsilateral seed ROIs e the corticospinal tract, CST (thick arrow), and both superior ROIs e transcallosal motor tracts, CCm (thin arrow), are depicted in 3D (E) and coronal 2D view (inset). (CS: central sulcus, pCS: pre-central sulcus, SFS: superior frontal sulcus). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Diffusion tensor imaging (DTI), a non-invasive MR imaging modality, has been shown to be potentially useful in differentiating PD from healthy controls [7e9]. In neural fiber tracts, water diffuses asymmetrically, being fastest along the major axis parallel to the direction of the fibers. The DTI indices of apparent diffusion coefficient (ADC) and fractional anisotropy (FA) describe the magnitude (diffusivity) and directionality (anisotropy) of diffusion of water molecules in the brain respectively [10,11]. Increased ADC and reduced FA values reveal underlying histopathological processes such as gliosis, demyelination, edema and axonal loss. Hence, they are used to elucidate the integrity of tissue microstructure in the brain. Fiber tracking, or tractography, uses the similarity in directionality of anisotropic diffusion between adjacent voxels on DTI to demonstrate neuronal projections from a single or connecting two functional center(s). Using two-region DTI tractography, the corticospinal and transcallosal fibers of the motor pathway can be localized and differentiated in the subcortical white matter. The objective of this case control DTI study, aided by two region tractography, is to determine if different anatomical brain regions representing the extrapyramidal and pyramidal motor systems,
transcallosal motor connection and corpus callosum are differentially affected in patients with PIGD parkinsonism compared to typical PD, and controls. 2. Methods 2.1. Study subjects All study subjects gave written informed consent. Approval from the institutional ethics committee was obtained for this study. Patients who presented initially with rest tremor, bradykinesia and rigidity and diagnosed with PD based on the United Kingdom PD Brain Bank clinical criteria [12], by a movement disorders neurologist in a tertiary referral center were included. Parkinsonian patients who presented initially with predominant postural instability, frequent falls, freezing and walking difficulty were defined as PIGD parkinsonism following criteria used in the literature [13]. Exclusion criteria included those with evidence of cognitive dysfunction (based on mini mental state score), severe joint problems or structural lower limb abnormalities, hemiplegia or focal neurological deficits to suggest a stroke, or neurological signs suggestive of normal pressure hydrocephalus, progressive supranuclear gaze palsy or multiple system atrophy. For every PD patient, we sought a PIGD patient and control of similar age and gender. The controls were healthy volunteers with no evidence of PD, recruited over the same period. A history of common vascular risk factors, such as hypertension, diabetes or hyperlipidemia, was not an exclusion criterion for all the study subjects.
L.-L. Chan et al. / Parkinsonism and Related Disorders 20 (2014) 53e59
55
Fig. 2. FA color maps depicting placement of circle ROIs representing the extrapyramidal, pyramidal corticospinal (CST) and corpus callosal (CC) tracts in the (A) pons CST, (B) substantia nigra and cerebral peduncle CST, (C) putamen, thalamus and internal capsule CST, (D) corpus callosum genu and splenium, (E) corona radiata CST and corpus callosum body (CCm) and (F) centrum semiovale CST and CCm. Identification of the CST and CC motor tracts at the corona radiata (E) and centrum semiovale (F) for ROI placement was aided by the tract volumes embedded on the b0 diffusion images (images on right). To evaluate the risk of falling, all study subjects were assessed using the Tinetti gait and balance test [14]. This comprised of a balance (9 items) and a gait (7 items) section. Each item was scored using a scale of 0, 1 or 2 and the total possible score was 28, with a lower score indicating more severe gait impairment. In addition, information on age, gender, age of onset of disease, family history, and severity of PD (Hoehn and Yahr scale, H&Y) were collected. A total of 100 study subjects were screened, of which 35 were excluded because they did not satisfy the inclusion and exclusion criteria or were unwilling to participate in study/unable to complete the MRI examination. 2.2. MR imaging The MR scans were performed on a 3 T Scanner (Siemens Trio, Erlangen, Germany), using a 12 channel phased array head coil. To reduce head motion, the subjects’ head was stabilized using Velcro straps, foam supports and padding, and patients were scanned during their “on” state. The scans were acquired in the axial plane, but tilted to the anterior commissureeposterior commissure (ACePC) line. The DTI scan was a spin-echo echo planar imaging (SE-EPI) sequence, using the following parameters: repetition time, TR ¼ 8200 ms; echo time, TE 86 ms; two b values (0 and 800 s/mm2); diffusion sensitization in 30 non-collinear directions; inplane resolution of 1.9 1.9 mm (FOV 240 240 mm, matrix 128 128); 64 contiguous 2 mm slices; and iPAT factor 2. The other pulse sequences were: (1) T2weighted (TR/TE 4530/84 ms, matrix 448 448), (2) MPRAGE with (TR/TE/TI 2200/ 3.0/900 ms), flip angle of 9 , matrix 256 256, and [3] fluid attenuated inversion recovery (FLAIR) with TR/TE/TI 7150/74/2535 ms and matrix 256 256. 2.3. DTI fiber tracking Three-dimensional (3D) reconstructions of fiber tracts were performed using the commercially available syngo DTI Tractography software (Siemens Healthcare Sector, Erlangen, Germany). The tractography algorithm is based on a deterministic streamline (FACT) approach and tracts are calculated with a Runge Kutta 4th order optimization. Tracking was terminated when the FA value dropped below 0.2 or if the curvature exceeded 30 between two contiguous eigen-vectors, and the step length was 1 mm [15,16]. Two seed regions of interest (ROIs) were drawn in each cerebral hemisphere on the 2-dimensional (2D) fractional anisotropy color maps (Fig. 1). The superior seed ROI was manually placed over the motor strip, based on anatomic sulcal morphology at the vertex and in correlation with the motor activation areas on fMRI. The inferior seed area was drawn in the blue portions of the ventral pons. Fiber tracts passing through ipsilateral superior and inferior seed ROIs descended through the known corticospinal tract (CST). Fiber tracts connecting the two superior seed ROIs in the motor strips traverse the body of the corpus callosum and were designated transcallosal motor tracts (CCm). 2.4. Image analysis FA maps with directions encoded in color (red (in web version): lefteright, blue (in web version): superioreinferior, green (in web version): anterioreposterior),
trace-ADC maps, trace-diffusion weighted images, and b0 diffusion images were displayed side by side. For the FA and ADC analysis, circle ROIs (Fig. 2) were drawn over various brain regions on the FA maps and ADC maps based on anatomical knowledge and also aided by superimposing the 3D CST and CCm tract volumes on structural b0 diffusion images (Fig. 2E,F). These ROIs yielded averaged values of FA and ADC. Anatomical brain regions (Fig. 2) representing the (i) extrapyramidal system: substantia nigra, putamen, and ventrolateral thalamus, (ii) pyramidal system comprising corticospinal tract (CST) at these levels: pons, cerebral peduncle, posterior limb of the internal capsule, corona radiata and centrum semiovale, and (iii) corpus callosal tracts (CC) at the genu, splenium, and transcallosal motor connection (CCm) at the corpus callosum body and centrum semiovale were studied using ROIs volumes of 52.8 mm3 for the pons, putamen and thalamus, and 22.1 mm3 for all other structures. ROIs were drawn within the core of each structure on representative sections at 32 mm (pons), 12 mm (substantia nigra and cerebral peduncles), þ6 mm (putamen, thalamus, posterior limb of internal capsule), þ12 mm (genu and splenium of corpus callosum), þ26 mm (corona radiata and corpus callosum body) and þ36 mm (centrum semiovale) from the section containing the ACePC. Presence of infarcts on MRI were assessed by our neuroradiologist (LL-C) based on conventional T1- and T2-weighted MR sequences. ROIs were not taken from any areas of infarcts and did not affect the DTI measurements.
3. Statistical analysis Statistical analyses were carried out using the SPSS software. Pearson’s chi square test was used to compare the categorical variables and Student’s t-test for the continuous variables. The mean of the right and left FA and ADC values of various anatomical regions was used in the analysis. A logistic regression to distinguish between PD and PIGD was carried out with the following independent variables: FA/ADC value for the various brain regions, patients’ demographics, Tinetti score, and MRI infarcts. These were investigated for entry in the model by stepwise variable selection method. The variable entry and exit criteria were p < 0.05 and p > 0.10 respectively. Box plot was performed to display the distribution of corpus callosum (body) FA and ADC values among patients with different Tinetti score. Receiver operating characteristics (ROC) analysis was carried out to evaluate the sensitivity and specificity of the DTI parameters as a screening and diagnostic tool for PD. The area under the curve (AUC) highlighted the discriminative property of the scale. All statistical tests were evaluated at two-sided and statistical significance was at p < 0.05.
56
L.-L. Chan et al. / Parkinsonism and Related Disorders 20 (2014) 53e59
4. Results
5. Discussion
A total of 65 subjects including 25 PIGD, 21 PD and 19 controls participated. All three groups were matched for age and gender and their demographics were summarized in Table 1. The mean Tinetti score in PIGD and PD was significantly lower than controls, and PIGD had the lowest score among the 3 groups (Table 1). For common vascular risk factors, PIGD had the lowest frequency of hypertension but the highest frequency for diabetes mellitus and heart disease and MRI evidence of vascular infarcts (Table 1). The mean FA and ADC values of each of the brain region in the 3 groups of subjects were tabulated in Tables 2 and 3 respectively. The FA of the substantia nigra was lower in PD and PIGD compared to controls but the difference was not significant. However, the ADC value was higher in PD vs control (p ¼ 0.004) and PIGD vs control (p ¼ 0.046), but difference between PD vs PIGD was not significant. The ADC values were also greater in the thalamus (p ¼ 0.002, p ¼ 0.003), putamen (p ¼ 0.176, p ¼ 0.03) and CST at the centrum semiovale (p ¼ 0.038, p ¼ 0.479) and corona radiata (p ¼ 0.027, p ¼ 0.072) in PIGD vs controls and in PIGD vs PD. The FA value of the corpus callosum body (p ¼ 0.047) and corpus callosum genu (p ¼ 0.013) was lower in PIGD vs controls (p ¼ 0.047), and lower in PIGD vs PD for the corpus callosum body (p ¼ 0.008). The ADC values were greater in the corpus callosum splenium (p ¼ 0.087, p ¼ 0.018), body (p ¼ 0.004, p ¼ 0.001) and CCm in the centrum semiovale (p ¼ 0.003, p ¼ 0.01) in PIGD vs controls and in PIGD vs PD. In the multivariate analysis, after adjustments were made for the effects of the FA and ADC values for the various brain regions, vascular infarcts on MRI, and Tinetti score, the FA (p ¼ 0.023) and ADC (p ¼ 0.001) values in the corpus callosum body (CCm), but not in the other brain regions, differentiated between PIGD and PD. For PIGD patients, the FA value was lower for those with Tinetti score less than or equal to 16 compared to those more than 16 (p ¼ 0.02) (Fig. 3). The ADC value was higher for those with Tinetti score less than or equal to 16 compared to those more than 16 (p ¼ 0.03) (Fig. 4). For ADC at the corpus callosum body, the AUC was at 82.3%, and at cut-off of 760 106 mm2/s, the sensitivity was 76% and specificity was 71.4% to differentiate PIGD from PD. For FA, the AUC was at modest at 70.3%, and at cut-off of 870 103, the sensitivity was 68% and specificity was 62.0%. There was no specific FA or ADC value which gave a 100% sensitivity and specificity.
While studies have suggested gait symptoms in PD are less levodopa responsive and amenable to DBS surgery [5,6] the exact pathophysiologic differences between PD and PIGD are still unclear. A recent MR DTI in vascular parkinsonism revealed frontal lobe white matter microstructural disruption [17]. In our case control DTI tractrographic study comparing PIGD patients with PD and controls, we demonstrated DTI abnormalities in the substantia nigra, thalamus, CST, corpus callosum tracts and CCm in PIGD compared to controls. The abnormality in the substantia nigra was compatible with the levodopa responsiveness of these patients. Though the mean ADC values were higher in the thalamus, putamen, corpus callosum and CCm in PIGD compared to PD, multivariate analysis revealed that only the FA and ADC values in the corpus callosum body differentiated PIGD from PD. A previous DTI-based tractography study in healthy individuals identified numerous vertical segments of the corpus callosum, containing fibers projecting into prefrontal, premotor (and supplementary motor), primary motor and sensory areas, parietal, temporal, and occipital cortical areas [18]. Transcallosal motor fiber bundles were found to cross the corpus callosum in posterior part of body and splenium. The anatomic projections delineated in this study are compatible with our study observations that corpus callosum body abnormalities differentiate patients with predominant postural instability and gait disorder compared to PD and healthy controls. Cortical projections to the premotor and supplemental motor areas, and primary motor and sensory areas arise from the body of the corpus callosum. Clinical, neurophysiologic and imaging studies have suggested that disruptions of the cortico-basal ganglia and basal gangliaebrainstem networks and dysfunction of sensorimotor integration processes can lead to gait disturbance [19]. Our study provides evidence that differential disruption of the callosal projections to these motor and sensory areas rather than primary abnormalities in the basal ganglia could explain why some parkinsonism patients developed postural instability and gait problems more than others. In the DATATOP cohort, PIGD symptoms were more common at onset in PD patients with a rapid rate of disease progression and reported greater subjective intellectual and motor impairment than the tremor onset PD [13]. The greater impairment in the corpus callosum in PIGD as highlighted in our study suggests either that the degree of disease progression is greater or mixed pathologies are present compared to the PD.
Table 1 Clinical features of PD, PIGD and controls. PD
Number Gender Age (years) Tinetti score Balance Gait Total H&Y Vascular risk factors Diabetes Hypertension Heart disease Cholesterol MRI vascular infarcts *p < 0.05.
PIGD
Controls
p Value PD vs controls
PIGD vs controls
PD vs PIGD
e e
e e
21 17 males 4 females 72.0 4.8 (62.0, 82.0)
25 18 males 7 females 73.3 6.2 (63.0, 87.0)
19 16 males 3 females 71.5 5.0 (62.0, 80.0)
e e 0.967
0.543
11.2 2.9 (4, 15) 8.6 1.9 (3, 11) 19.9 4.4 (7, 24) 2.2 0.5 (2, 4)
9.3 3.4 (3, 16) 6.8 2.5 (2, 12) 16.1 5.3 (7, 26) 3.0 1.0 (2, 4)
15.3 0.7 (13, 16) 11.4 0.8 (10, 12) 26.7 1.2 (24, 28) e
<0.01* <0.01* <0.01*
<0.01* <0.01* <0.01*
5 (23.8%) 12 (57.1%) 3 (14.3%) 9 (42.9%) 4 (19.0%)
10 (40%) 13 (52%) 7 (28%) 11 (44%) 15 (60%)
3 (15.8%) 12 (63.2%) 1 (5.3%) 6 (31.6%) 5 (26.3%)
0.527 0.962 0.342 0.462 0.587
0.081 0.697 0.053 0.402 0.028*
0.690 0.047* 0.010* 0.010*
0.243 0.727 0.261 0.938 0.005*
L.-L. Chan et al. / Parkinsonism and Related Disorders 20 (2014) 53e59
57
Table 2 FA values of study subjects. ROI (103)
PD mean SD (range)
Extrapyramidal system Substantia nigra 519.1 101.1 (328.4, Thalamus 328.5 42.9 (239.6, Putamen 268.2 63.1 (122.6, Pyramidal system e corticospinal tract (CST) Pons 578.5 42.7 (504.5, Cerebral peduncle 714.7 52.3 (620.7, Internal capsule 763.7 36.9 (712.7, Corona radiata 558.2 59.5 (426.3, Centrum semiovale 413.2 76.3 (256.3, Corpus callosum tracts Genu 859.6 40.3 (757.0, Splenium 918.7 33.9 (849.3, Body (CCm) 875.9 60.4 (707.7, Centrum semiovale 480.0 84.3 (368.3, (CCm)
PIGD mean SD (range)
752.0) 384.9) 371.0)
521.8 80.6 (375.2, 667.5) 306.4 45.5 (209.6, 393.6) 260.14 67.5 (157.7, 384.0)
669.8) 839.8) 847.9) 694.0) 558.9) 930.7) 984.5) 949) 629.4)
Controls mean SD (range)
575.7 714.3 768.0 576.0 433.5
76.0 50.8 38.8 69.8 86.5
818.8 102.4 865.8 120.1 814.3 71.2 462.9 117.3
(428.2, (561.3, (698.1, (441.1, (260.9,
774.0) 805.8) 838.5) 735.4) 645)
(477.5, 913.3) (393, 977) (674, 940.5) (214.1, 795.2)
The Tinetti scale is a sensitive, reliable and valid clinical instrument for assessing the risk of falling and is a good measure of balance and gait in the elderly. It has also been shown to correlate with gait and postural stability UPDRS scores in PD patients [14]. PIGD patients are at a very high risk for falls and thus their Tinetti scores will be particularly relevant as a measure of the clinical severity. In our study, there was a lower FA value and a higher ADC value (corpus callosum body) in PIGD patients with low Tinetti score compared to those with a high score (Figs. 3 and 4). This provides further supporting evidence that the corpus callosum abnormality is associated with gait dysfunction. While the correlation of a reduced FA value or increased ADC value with degeneration of specific pathological structures (active neuronal loss, gliosis or inflammation) is not entirely clear, these changes in the corpus callosum probably represent a combination of microstructural damage and cortical atrophy with secondary transcallosal degeneration. In PD, the substantia nigra is the most severely affected primary site of pathology and secondary degeneration could be present in the basal ganglia and extra-striatal regions in the later stages. In a clinico-pathologic study, the majority of tremor onset PD cases had Lewy body pathology while the majority of PIGD-onset cases had other variable forms of pathology [2]. The differences in the ADC and FA values in the substantia nigra were greater in PD vs controls
PD vs controls
PIGD vs controls
PD vs PIGD
0.780 0.784 0.978
0.824 0.678 0.812
0.993 0.262 0.908
634.7) 765.2) 819.8) 747.4) 531.8)
0.891 0.531 0.393 0.922 0.966
0.948 0.523 0.203 0.878 0.818
0.984 0.999 0.927 0.636 0.648
(806.3, 976) (836.3, 963) (771, 959.3) (320.8, 822.8)
0.567 0.878 0.123 0.998
0.013* 0.218 0.047* 0.825
0.114 0.07 0.008* 0.849
536.7 58.3 (436.5, 678.2) 318.5 53.2 (239.4, 456.9) 272.3 62.9 (131.1, 411.0) 570.5 698.6 747.4 566.2 419.3
27.5 34.3 41.6 67.9 64.4
882.8 45.3 906.5 33.1 857.0 46.7 482.0 111.1
p Value
(523.8, (651.0, (690.9, (464.4, (314.7,
compared to PIGD vs controls, and the absolute values of FA and ADC were lower and higher in PD compared to PIGD, suggesting more severe involvement of the substantia nigra in PD compared to PIGD. Our study showed that vascular infarcts were more commonly seen in PIGD compared to PD and controls. It is possible that vascular lesions contribute to the pathophysiology and the clinical phenotype of patients with PIGD may overlap with “vascular” form of parkinsonism. Specific neuronal structures or tracks could have been dysregulated as a consequence of vascular damage. Recent structural MR studies have also showed a higher prevalence of leukoaraiosis in PIGD compared to PD. A previous structural MRI study and meta-analysis of published literature has examined corpus callosum as a marker of cortical pathology in neurodegenerative diseases [20] and concluded that PD patients do not have callosal atrophy compared to other neurodegenerative diseases. However, it unclear if any of these patients belonged to the PIGD phenotype. In a recent study involving a general elderly population, DTI was used to examine the genu and splenium of the corpus callosum only, an independent association of gait with genu abnormalities and a correlation of the DTI abnormality with the validated Tinetti gait scores was found [1]. DTI abnormalities in the corpus callosum have been shown to be present in corticobasal syndrome [21,22] and the anterior corpus callosum is more affected
Table 3 ADC values of study subjects. ROI (106 mm2/s)
PD mean SD (range)
Extrapyramidal system Substantia nigra 901.6 148.4 Thalamus 749.1 67.9 Putamen 793.4 129.7 Pyramidal system e corticospinal tract (CST) Pons 798.8 57.8 Cerebral peduncle 942.5 123.4 Internal capsule 732.2 38.9 Corona radiata 755.6 40.4 Centrum semiovale 816.0 66.6 Corpus callosum tracts Genu 759.2 72.6 Splenium 671.4 86.7 Body (CCm) 717.8 77.9 Centrum semiovale (CCm) 847.7 82.5 *p < 0.05.
(564.7, 1164.5) (649.8, 928.7) (651.4, 1207.8) (677.9, (666.5, (643.1, (685.8, (679.9,
909.1) 1173.8) 788.4) 828.5) 942.7)
(553.0, (504.8, (586.7, (715.9,
905.0) 856) 894.7) 1076.7)
PIGD mean SD (range)
856.1 116.8 (569.5, 1039.3) 841.3 122.2 (678.8, 1140.3) 986.9 353.7 (480.4, 2239) 785.6 62.2 1000.3 164.9 741.3 59.1 802.9 88.7 844.9 98.1 859.2 782.0 876.9 961.3
280.0 179.6 134.6 178.7
(660.6, (762.1, (631.3, (636.4, (692.2,
884.7) 1588.4) 864.5) 981.4) 1063.0)
(99.8, 1727.5) (565, 1437.7) (665, 1192) (648.0, 1326.5)
Controls mean SD (range)
p Value PD vs controls
PIGD vs controls
PD vs PIGD
746.0 182.3 (105.8, 1005.5) 741.4 57.8 (636.4, 866.0) 849.0 176.8 (541.9, 1267.8)
0.004* 0.961 0.765
0.046* 0.002* 0.176
0.558 0.003* 0.030*
773.8 43.4 918.2 203.8 728.2 36.0 745.4 72.2 780.6 80.8
0.341 0.889 0.960 0.893 0.383
0.769 0.242 0.630 0.027* 0.038*
0.705 0.472 0.791 0.072 0.479
0.876 0.840 0.096 0.860
0.066 0.087 0.004* 0.003*
0.171 0.018* 0.001* 0.010*
730.4 695.1 766.6 826.5
89.8 95.4 88.5 77.0
(669.6, (200.1, (672.2, (653.3, (692.3,
841.7) 1176.5) 815.8) 956.9) 1021.9)
(547.3, 874.3) (546.7, 857.7) (660, 993.7) (715.7, 954.7)
58
L.-L. Chan et al. / Parkinsonism and Related Disorders 20 (2014) 53e59
clinical trials or for prognostic purposes. Second; there is no gold standard validation of the diagnosis of PIGD and it is possible that this entity is clinically heterogeneous even though gait abnormality is the predominant manifestation. Hence there may be overlap with “vascular” parkinsonism or even some of the atypical parkinsonian disorders. Third; as there is paucity of white matter tracts in the deep gray matter structures, the FA values in these structures in our study have to be interpreted with caution. In conclusion, we demonstrated for the first time that DTI abnormalities along the transcallosal motor tract in the body of the corpus callosum, but not the substantia nigra, differentiated PIGD from PD, and the degree of corpus callosum body abnormality correlated with the Tinetti score (a measure of risk of falls). Author contributions
Fig. 3. For PIGD patients, the FA value was lower for those with Tinetti score less than or equal to 16 compared to those more than 16 (p ¼ 0.02).
in progressive supranuclear palsy compared to PD [23]. These associations bear some similarities to PIGD. As PIGD may have varied pathologies and clinical associations, it is conceivable that the pathophysiologic process may overlap with some of the atypical parkinsonian syndromes. To assess the potential utility of the FA and ADC values in clinical practice, we conducted a ROC analysis to assess the sensitivity and specificity of these values. Ideal screening and diagnostic tests should have both high positive and negative predictive values. As expected, there was a considerable overlap of the FA and ADC values (corpus callosum) between PD and PIGD. This is to be expected because of the considerable overlap in gait dysfunction. For example, the area under curve for the ADC (corpus callosum body) was 82.3%, and with a cut-off value of 760 106 mm2/s the sensitivity was 76% and specificity 71.4%. While these values cannot be used for screening, their clinical utility should be addressed in a long term longitudinal study. Our study has limitations. First; the cross-section design did not allow measurement of FA or ADC values serially for individual subjects in a longitudinal manner. A prospective study to measure the DTI values in the corpus callosum would be useful to monitor the rate of decline in these patients. This might provide corroborative data when monitoring patients’ response to neuroprotective
Fig. 4. For PIGD patients, the ADC value was higher for those with Tinetti score less than or equal to 16 compared to those more than 16 (p ¼ 0.03).
EK Tan, LL Chan designed the study. KM Ng, H Rumpel, EK Tan, LL Chan obtained the data. HH Li, S Fook-Chong, KM Ng performed the statistical analysis. LL Chan and EK Tan wrote the manuscript. All the authors were involved in the interpretation of the data and revision of the manuscript. LL Chan supervised the study. Financial disclosures All the authors have no conflict of interest and financial disclosures related to this study to declare. The corresponding author declares that the Authors take full responsibility for the data, the analyses and interpretation, and the conduct of the research; that the Author has full access to all of the data; and that the Author has the right to publish any and all data, separate and apart from the guidance of any sponsor. This work is supported by the National Medical Research Council, Duke-NUS Graduate Medical School, Singapore Millennium Foundation, SingHealth Foundation. Acknowledgments We thank the National Medical Research Council, Singapore Millennium Foundation, SingHealth Foundation, and Duke NUS Graduate Medical School for their support. References [1] Bhadelia RA, Price LL, Tedesco KL, Scott T, Qiu WQ, Patz S, et al. Diffusion tensor imaging, white matter lesions, the corpus callosum, and gait in the elderly. Stroke 2009;40:3816e20. [2] Rajput AH, Pahwa R, Pahwa P, Rajput A. Prognostic significance of the onset mode in parkinsonism. Neurology 1993;43:829e30. [3] Lee SJ, Kim JS, Lee KS, An JY, Kim W, Kim YI, et al. The severity of leukoaraiosis correlates with the clinical phenotype of Parkinson’s disease. Arch Gerontol Geriatr 2009;49:255e9. [4] Alves G, Larsen JP, Emre M, Wentzel-Larsen T, Aarsland D. Changes in motor subtype and risk for incident dementia in Parkinson’s disease. Mov Disord 2006;21:1123e30. [5] Jankovic J, Aguilar LG. Current approaches to the treatment of Parkinson’s disease. Neuropsychiatr Dis Treat 2008;4:743e57. [6] Lang AE, Lozano AM. Parkinson’s disease. First of two parts. N Engl J Med 1998;339:1044e53. [7] Vaillancourt DE, Spraker MB, Prodoehl J, Vaillancourt DE, Spraker MB, Prodoehl J, et al. High-resolution diffusion tensor imaging in the substantia nigra of de novo Parkinson disease. Neurology 2009;72:1378e84. [8] Gattellaro G, Minati L, Grisoli M, Mariani C, Carella F, Osio M, et al. White matter involvement in idiopathic Parkinson disease: a diffusion tensor imaging study. AJNR Am J Neuroradiol 2009;30:1222e6. [9] Chan LL, Rumpel H, Yap K, Lee E, Loo HV, Ho GL, et al. Case control study of diffusion tensor imaging in Parkinson’s disease. J Neurol Neurosurg Psychiatry 2007;78:1383e6. [10] Basser PJ, Pierpaoli C. Microstructural and physiological features of tissues elucidated by quantitative-diffusion-tensor MRI. J Magn Reson B 1996;111: 209e19.
L.-L. Chan et al. / Parkinsonism and Related Disorders 20 (2014) 53e59 [11] Beaulieu C. The basis of anisotropic water diffusion in the nervous system e a technical review. NMR Biomed 2002;15:435e55. [12] 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):181e4. [13] Jankovic J, McDermott M, Carter J, Gauthier S, Goetz C, Golbe L, et al. Variable expression of Parkinson’s disease: a base-line analysis of the DATATOP cohort. The Parkinson Study Group. Neurology 1990;40:1529e34. [14] Kegelmeyer DA, Kloos AD, Thomas KM, Kostyk SK. Reliability and validity of the Tinetti mobility test for individuals with Parkinson disease. Phys Ther 2007;87:1369e78. [15] Wakana S, Jiang H, Nagae-Poetscher LM, van Zijl PC, Mori S. Fiber tract based atlas of human white matter anatomy. Radiology 2004;230:77e87. [16] Nucifora PG, Verma R, Lee SK, Melhem ER. Diffusion-tensor MR imaging and tractography: exploring brain microstructure and connectivity. Radiology 2007;245:367e84. [17] Wang HC, Hsu JL, Leemans A. Diffusion tensor imaging of vascular parkinsonism: structural changes in cerebral white matter and the association with clinical severity. Arch Neurol 2012:1e9 [doi].
59
[18] Hofer S, Frahm J. Topography of the human corpus callosum revisitede comprehensive fiber tractography using diffusion tensor magnetic resonance imaging. Neuroimage 2006;32:989e94. [19] Takakusaki K, Tomita N, Yano M. Substrates for normal gait and pathophysiology of gait disturbances with respect to the basal ganglia dysfunction. J Neurol 2008;255:19e29. [20] Wiltshire K, Foster S, Kaye JA, Small BJ, Camicioli R. Corpus callosum in neurodegenerative diseases: findings in Parkinson’s disease. Dement Geriatr Cogn Disord 2005;20:345e51. [21] Boelmans K, Bodammer NC, Suchorska B, Kaufmann J, Ebersbach G, Heinze HJ, et al. Diffusion tensor imaging of the corpus callosum differentiates corticobasal syndrome from Parkinson’s disease. Parkinsonism Relat Disord 2010;16: 498e502. [22] Boelmans K, Kaufmann J, Bodammer N, Ebersbach G, Behlau G, Heinze HJ, et al. Involvement of motor pathways in corticobasal syndrome detected by diffusion tensor tractography. Mov Disord 2009;24:168e75. [23] Ito S, Makino T, Shirai W, Hattori T. Diffusion tensor analysis of corpus callosum in progressive supranuclear palsy. Neuroradiology 2008;50:981e5.