| | Disrupted thalamic prefrontal pathways in patients with idiopathic dystoniaReceived 12 October 2007; received in revised form 5 December 2007; accepted 25 January 2008. Abstract There are quantifiable abnormalities in water diffusion properties of the white matter in thalamic and prefrontal areas in patients with idiopathic dystonia (ID). However, it is unclear which pathways are disrupted in these patients. Using probabilistic tractography of high resolution DTI, we reconstructed thalamic prefrontal pathways in seven patients with ID and seven matched controls. Resulting fibers were registered onto the stereotaxic space and submitted to a voxel-wise statistical analysis comparing patients and controls. Patients with ID exhibited less thalamic prefrontal connections, particularly involving fibers traveling from the thalamus to the middle frontal gyrus. These results corroborate neurophysiologic findings of reduced and asynchronous thalamic prefrontal input, and emphasize the structural correlates of the pathophysiology of ID. 1. Introduction  Idiopathic dystonia (ID) has long been considered a functional brain disorder, as routine neuroradiological studies of patients with ID are usually normal [1]. The concept of a functional disorder is also supported by research employing Positron Emission Tomography (PET), which demonstrates a consistent pattern of abnormal metabolism in patients with ID [2], [3]. These PET studies show an enhanced glucose uptake involving a cortico-striatal network, affecting the posterior putamen, globus pallidus, cerebellum and the supplementary motor area in subjects with ID [2], [4], [5]. More recent studies using high resolution structural magnetic resonance imaging (MRI) indicate that ID is not only a functional disorder, but is also associated with differences in brain structure. Manual assessment of basal ganglia volume shows that the putamen is larger in patients with cranial and limb dystonia [6]. Similarly, voxel based morphometry studies demonstrated increased gray matter in the motor cortex, cerebellar flocculus and right globus pallidus internus in cervical dystonia [1], increased gray matter in the putamen in blepharospasm [7], and increased gray matter in the primary somatosensory cortex in focal hand dystonia [8]. White matter abnormalities have also been found in ID. Studies using diffusion tensor imaging (DTI) demonstrated that patients with ID show disruption of water flow within the white matter, suggesting that impaired connectivity is potentially an intrinsic feature of the pathophysiology of ID. From DTI images, it is possible to estimate the mean diffusivity of water molecules (MD) and the fractional anisotropy of molecular diffusion (FA). Carbon et al. observed that patients with generalized dystonia and DYT1 mutation show differences in FA underlying the sensorimotor cortex [9], and our group demonstrated FA and MD abnormalities in the thalamic and prefrontal regions in patients with focal or generalized sporadic ID [10]. These findings suggest that white matter fibers guiding the synchronized diffusion of water are abnormality arranged in ID. However, a significant limitation of studies employing the statistical analysis of FA and MD is that conclusions are limited to the general anatomic localization of abnormalities, without information about which fibers are in fact abnormal. An inference is usually made based on the approximate location of the abnormalities, but still without the certainty of a particular pathway or direction of flow. Based on previous DTI findings from our group and from Carbon et al. [9], [10] showing abnormalities involving the frontal lobe and white matter surrounding the thalamus, we aimed to investigate in this study whether thalamic prefrontal pathways are abnormal in patients with ID. We employed probabilistic tractography of DTI images to reconstruct the white matter pathways from the thalamus to the prefrontal cortex and compared the connectivity distribution between patients with ID and matched controls. 2. Subjects and methods 2.1. Subjects We studied seven subjects with ID (six women, mean age =58±18y). All patients had cervical dystonia as their most prominent symptom (one patient had concurrent spasmodic dysphonia, and two patients had cervical dystonia in the context of a generalized dystonia and were DYT1 negative). Three patients had right laterocollis, while three had left laterocollis and one had anterocollis. We also studied seven age and gender “person-to-person” matched healthy controls (seven women, mean age=58±18y, no age difference between groups: paired t(6)=2.09, p=0.08). All subjects signed a written informed consent to participate in the study, which was approved by the institutional review board of the Medical University of South Carolina. The same patients were also studied in our previous manuscript evaluating DTI abnormalities in ID [10]. All subjects were scanned in a Philips 3T Intera scanner (Best, Netherlands) using an eight-channel head coil at the Medical University of South Carolina. Routine DTI images were acquired using single shot EPI with cubic voxels of 2.5×2.5×2.5mm, employing the default Philips 15 diffusion directions with b value of 1000s/mm2 (time of acquisition: 5:40min). DTI images were transformed into Nifti format, whilst variations in acquisition geometry and gradients were computed using DtoA (Paul Morgan, http://www.nottingham.ac.uk/radiology/paul/software/). FSL's (FMRIB's Software Library, www.fmrib.ox.ac.uk/fsl) Diffusion Toolkit (FDT) was then used to pre-process the diffusion weighted images and construct DTI data. Images underwent eddy current correction through affine transform of each diffusion weighted image (DWI) to the base b=0 T2 weighted image. Eddy current correction aims to remove spatial distortion in the DWIs due to application of diffusion gradients in various directions, and ensures that all diffusion images acquired of the same slice are in alignment and so a pixel-wise calculation of the diffusion tensor may be performed. The first volume (which is effectively a T2 weighted spin-echo echo planar image) of the DTI acquisition was extracted, and the pixel-wise calculation of the diffusion tensor was performed using FSL's DTIFit, including application of a binary brain mask extracted using FSL's Brain Extraction Tool (BET) with fractional threshold of 0.3 to prevent erroneous DTI calculation in the noise background outside the head. Next, FSL's Bedpost (Bayesian Estimation of Diffusion Parameters Obtained using Sampling Techniques) was applied to the data. Bedpost runs Markov Chain Monte Carlo sampling to build up distributions on diffusion parameters at each voxel. It creates all the files necessary for running probabilistic tractography. Finally, probabilistic tractography was estimated using FSL's probtrack. Probabilistic tractography was performed using a seed mask and waypoint masks in order to generate a connectivity distribution from voxels in the seed mask retaining only those paths that pass through the waypoint mask. The seed mask was located in the thalamus and the waypoint mask was located in the middle frontal gyrus. Both masks were chosen in order to match the locations where our group observed significant differences in FA and MD in our previous study [10]. In order to avoid a post hoc bias, we did not employ masks corresponding to the suprathreshold voxels from our previous analyses. Instead, we chose to employ masks defined by an independent anatomical atlas. Both masks were obtained from the Anatomical Automatic Labeling dataset (specified in Table 1) (http://www.cyceron.fr/freeware/), and were linearly transformed into each subject's native space, where probtrack was run. The resulting probabilistic tractography images from each subject were linearly transformed into standard stereotaxic MNI space (Fig. 1) and smoothed using an Isotropic Gaussian Kernel of 4mm. | | |  | | Center of mass | Range (min–max) | Volume (mm3) | |
|---|
| x | y | z | x | y | z | |
|---|
| Mask | | | Left thalamus | −11 | −19 | 7 | −22, 0 | −34, −6 | −2, 18 | 8800 | | | Right thalamus | 13 | 19 | 7 | 2, 24 | −34, −6 | −2, 18 | 8456 | | |
| | | Way points | | | Left middle frontal region | −34 | 32 | 34 | −54, −18 | −14, 66 | −2, 64 | 38904 | | | Right middle frontal region | 37 | 32 | 33 | 20, 58 | −12, 64 | −2, 64 | 40832 | | | | |
Statistical analyses were applied to spatially normalized smoothed probalistic tractography maps using a paired t-test embedded into the software NPM [11]. Results were corrected for multiple comparisons using a False Discovery Rate (FDR) corrected threshold of p<0.05. We also explored the possibility of difference between patients with right sided laterocollis (n=3) versus left sided laterocollis (n=3). This analysis involved a non-paired t-test with a False Discovery Rate (FDR) corrected threshold of p<0.05. 3. Results We observed a significant reduction in the probability of tracks from the thalamus to the frontal lobe in the right and left hemisphere of patients with ID (Fig. 2). In particular, fibers were less frequently encountered in the pathway projecting from the thalamus, through the white matter between the caudate and the putamen, to the anterior portion of the frontal lobe. There was not a significant increase in fiber probability in the group of patients with ID. We did not observe significant differences between patients with left and right sided laterocollis, possibly due to the reduced number of patients in each group (n=3). 4. Discussion Our results confirm that patients with ID show an abnormal pattern of white matter connectivity. Specifically, we demonstrated that thalamic prefrontal pathways are disrupted in ID. These findings support current evidence that ID is not only a functional disorder, but it is also associated with structural brain changes. Combined with evidence from prior functional and structural studies, the observation of a definite pathway abnormality in ID tightens the link between the functional and structural nature of its pathophysiology. Even though the mechanisms underlying the generation of dystonic movements are not completely understood, it is possible that disrupted connections within the basal nuclei play a central role. In particular, electrophysiological studies suggest that the thalamo-cortical input in ID is reduced during rest and asynchronous during the execution of a dystonic movement [12]. While this can be a consequence of intrinsic abnormal neuronal firing within the thalamus, it might also be related to irregular connectivity between the thalamus and the prefrontal cortex. Based on the results from this study of probabilistic tractography of high resolution DTI, we suggest that patients with ID exhibit a quantifiable disruption of the thalamo-prefrontal pathways, which might play an important role in the generation of abnormal dystonic movements. References [1]. [1]Draganski B, Thun-Hohenstein C, Bogdahn U, Winkler J, May A. “Motor circuit” gray matter changes in idiopathic cervical dystonia. Neurology. 2003;61:1228–1231. [2]. [2]Carbon M, Trost M, Ghilardi MF, Eidelberg D. Abnormal brain networks in primary torsion dystonia. Adv Neurol. 2004;94:155–161. MEDLINE [3]. [3]Eidelberg D, Moeller JR, Antonini A, Kazumata K, Nakamura T, Dhawan V, et al. Functional brain networks in DYT1 dystonia. Ann Neurol. 1998;44:303–312. MEDLINE |
CrossRef
[4]. [4]Eidelberg D. Brain networks and clinical penetrance: lessons from hyperkinetic movement disorders. Curr Opin Neurol. 2003;16:471–474. MEDLINE |
CrossRef
[5]. [5]Galardi G, Perani D, Grassi F, Bressi S, Amadio S, Antoni M, et al. Basal ganglia and thalamo-cortical hypermetabolism in patients with spasmodic torticollis. Acta Neurol Scand. 1996;94:172–176. MEDLINE |
CrossRef
[6]. [6]Black KJ, Ongur D, Perlmutter JS. Putamen volume in idiopathic focal dystonia. Neurology. 1998;51:819–824. MEDLINE [7]. [7]Etgen T, Mühlau M, Gaser C, Sander D. Bilateral grey-matter increase in the putamen in primary blepharospasm. J Neurol Neurosurg Psychiatry. 2006;77:1017–1020.
CrossRef
[8]. [8]Garraux G, Bauer A, Hanakawa T, Wu T, Kansaku K, Hallett M. Changes in brain anatomy in focal hand dystonia. Ann Neurol. 2004;55:736–739. MEDLINE |
CrossRef
[9]. [9]Carbon M, Kingsley PB, Su S, Smith GS, Spetsieris P, Bressman S, et al. Microstructural white matter changes in carriers of the DYT1 gene mutation. Ann Neurol. 2004;56:283–286. MEDLINE |
CrossRef
[10]. [10]Bonilha L, de Vries PM, Vincent DJ, Rorden C, Morgan PS, Hurd MW, et al. Structural white matter abnormalities in patients with idiopathic dystonia. Mov Disord. 2007;22:1110–1116.
CrossRef
[11]. [11]Rorden C, Bonilha L, Nichols TE. Rank-order versus mean based statistics for neuroimaging. Neuroimage. 2007;35:1531–1537. MEDLINE |
CrossRef
[12]. [12]Vitek JL. Pathophysiology of dystonia: a neuronal model. Mov Disord. 2002;17(Suppl. 3):S49–S62.
CrossRef
a Department of Neurology, Medical University of South Carolina, Charleston, SC, USA b Murray Center for Research on Parkinson's Disease and Related Disorders, Medical University of South Carolina, Charleston, SC, USA c Department of Neurology, Neuroimaging Center, Institute of Behavioral and Cognitive Neuroscience, University of Groningen, The Netherlands d Department of Radiology, Medical University of South Carolina, Charleston, SC, USA e Department of Communication Sciences and Disorders, University of South Carolina, SC, USA f Department of Physics and Astronomy, University of South Carolina, SC, USA  Corresponding author. Medical University of South Carolina, 326 Calhoun Street, Suite 308, Charleston, SC 29425, USA. Tel.: +1 843 792 3222; fax: +1 843 792 8626.
PII: S1353-8020(08)00045-X doi:10.1016/j.parkreldis.2008.01.018 © 2008 Elsevier Ltd. All rights reserved. | |
|
FULL TEXT
