J Neurol (2014) 261:291–299 DOI 10.1007/s00415-013-7186-6
ORIGINAL COMMUNICATION
White matter damage is related to ataxia severity in SCA3 J.-S. Kang • J. C. Klein • S. Baudrexel R. Deichmann • D. Nolte • R. Hilker
•
Received: 5 September 2013 / Revised: 5 November 2013 / Accepted: 5 November 2013 / Published online: 23 November 2013 Ó Springer-Verlag Berlin Heidelberg 2013
Abstract Spinocerebellar ataxia type 3 (SCA3) is the most frequent inherited cerebellar ataxia in Europe, the US and Japan, leading to disability and death through motor complications. Although the affected protein ataxin-3 is found ubiquitously in the brain, grey matter atrophy is predominant in the cerebellum and the brainstem. White matter pathology is generally less severe and thought to occur in the brainstem, spinal cord, and cerebellar white matter. Here, we investigated both grey and white matter pathology in a group of 12 SCA3 patients and matched controls. We used voxel-based morphometry for analysis of tissue loss, and tract-based spatial statistics (TBSS) on diffusion magnetic resonance imaging to investigate microstructural pathology. We analysed correlations between microstructural properties of the brain and ataxia severity, as measured by the Scale for the Assessment and Rating of Ataxia (SARA) score. SCA3 patients exhibited significant loss of both grey and white matter in the cerebellar hemispheres, brainstem including pons and in lateral
thalamus. On between-group analysis, TBSS detected widespread microstructural white matter pathology in the cerebellum, brainstem, and bilaterally in thalamus and the cerebral hemispheres. Furthermore, fractional anisotropy in a white matter network comprising frontal, thalamic, brainstem and left cerebellar white matter strongly and negatively correlated with SARA ataxia scores. Tractography identified the thalamic white matter thus implicated as belonging to ventrolateral thalamus. Disruption of white matter integrity in patients suffering from SCA3 is more widespread than previously thought. Moreover, our data provide evidence that microstructural white matter changes in SCA3 are strongly related to the clinical severity of ataxia symptoms. Keywords Spinocerebellar ataxia 3 MachadoJoseph disease DTI Diffusion Ataxia
Introduction J.S. Kang and J.C. Klein equally contributed to this work.
Electronic supplementary material The online version of this article (doi:10.1007/s00415-013-7186-6) contains supplementary material, which is available to authorized users. J.-S. Kang J. C. Klein S. Baudrexel R. Hilker Department of Neurology, Goethe-University of Frankfurt, Schleusenweg 2-16, 60528 Frankfurt am Main, Germany J.-S. Kang J. C. Klein (&) S. Baudrexel R. Deichmann Brain Imaging Center (BIC), Goethe-University of Frankfurt, Schleusenweg 2-16, 60528 Frankfurt am Main, Germany e-mail: [email protected] D. Nolte Institute of Human Genetics, University of Giessen, Schlangenzahl 14, 35392 Giessen, Germany
Spinocerebellar ataxia type 3 (SCA3), also known as Machado-Joseph disease, is the most frequent inherited cerebellar ataxia in Europe, the US and Japan [1]. SCA3 is an autosomal-dominant neurodegenerative disease with a variable clinical presentation, comprising ataxia, spasticity, parkinsonism, neuropathy, dystonia, and ophthalmoplegia [1]. SCA3 is caused by an abnormal CAG trinucleotide expansion on chromosome 14 (gene locus 14q32.12), encoding for polyglutamine in the abnormal gene product ataxin-3. The repeat length of the pathologically expanded allele varies between 45 and 51 repeats with reduced penetrance, and 52 and 86 copies with full penetrance. However, the exact pathogenesis of SCA3 is currently
123
292
unknown. A gain of function of the ‘‘toxic protein’’ is one possible pathomechanism discussed [2]. Neuronal intranuclear inclusion bodies are a pathological hallmark at autopsy, and may also play a role in the pathogenesis of SCA3 [3, 4]. A greater repeat length seems to precipitate an earlier disease onset [5]. Although ataxin-3 is found ubiquitously in the brain, macroscopic brain atrophy has been shown predominantly in the cerebellum and in the brainstem [6]. Neuroimaging studies, mostly employing standard magnetic resonance imaging (MRI), revealed marked cerebellar and pontine atrophy in SCA3 patients [7, 8]. The amount of brain atrophy correlated with the disease severity, and weakly with CAG repeat length [9]. Recently, evidence for cortical involvement has been reported in a longitudinal imaging study measuring cortical thickness [10]. In contrast, white matter pathology was generally considered less severe and, on histology, thought to be largely limited to brainstem, spinal cord, and cerebellar white matter [6]. However, one recent diffusion-weighted MRI (DWI) study demonstrated increased water diffusivity in supratentorial white matter [11]. In this observational study, we aimed to investigate both grey and white matter pathology in a group of 12 subjects with genetically proven SCA3. For this, we performed voxel-based morphometry on optimised anatomical MRI, and tract-based spatial statistics (TBSS) as well as tractography on DWI. Measures derived from the diffusion tensor model, such as fractional anisotropy (FA) and mean diffusivity (MD), have been shown to relate to the severity of degenerative and inflammatory diseases of the brain [12–14]. Therefore, we set out to explore white matter dysintegrity in SCA3 and, in particular, its relationship with clinically evident motor impairment.
Patients and methods Subjects Twelve patients with a molecular diagnosis of SCA3 participated in this study (five women, seven men; 50.5 ± 10.4 years, range 30–63 years). The mean CAG repeat length was 70.6 ± 3.8, range 65–78 copies. Mean disease duration was 11.0 ± 6.4 years and the mean Scale for the Assessment and Rating of Ataxia (SARA) [15] score amounted to 10.3 ± 4.6. Twelve age- and sex-matched control subjects (five women, seven men; 47.8 ± 11.02 years, ranging from 26 to 61 years) participated as a control group. None of the participants had a history of other neurological or psychiatric disease, and neurological examination was normal in the control group. All participants gave their informed consent before entering
123
J Neurol (2014) 261:291–299
the study. The study conformed to the latest revision of the declaration of Helsinki of 1995 and was approved by the ethics committee of the University Hospital Frankfurt. MR imaging All subjects underwent MRI on a Siemens Trio 3T (Siemens, Erlangen, Germany) scanner at Frankfurt University’s Brain Imaging Center. For all MR procedures, the head was immobilised using self-expanding foam cushions. Image acquisition used an 8-channel array head coil for signal reception, and the body coil for transmission. Anatomical images were recorded using T1-weighted spoiled FLASH (TE 2.4 ms, TR 7.6 ms, flip angle 18°, 176 sagittal slices, matrix 224 9 256, isotropic voxel size 1 mm, 2 repeats) [16]. Diffusion was measured along 60 isotropically distributed directions using spin-echo EPI [17] and GRAPPA [18] (SE-EPI, TE 95 ms, TR 9.3 s, 70 axial slices, matrix 104 9 104, isotropic voxel size 2 mm, acceleration factor 2, 33 reference lines, b value 1,000 s mm-2) and 10 non-diffusion weighted reference images. Diffusion scanning was repeated three times to increase SNR. Common space All data processing used tools from the FMRIB software library (FSL) [19]. To enable between-subject statistics, all images were non-linearly registered to a high-resolution MR template in MNI152 space provided with FSL. Parameters taking individual images into standard space and vice versa were recorded to enable analysis across individual imaging spaces. Data processing and analysis: VBM grey and white matter First, the two FLASH volumes were registered linearly to correct for potential subject movement. These volumes were then summed to reduce noise while maintaining high grey to white matter contrast. Anatomical data were then analysed with FSL-VBM [20]. First, images were brain-extracted and grey and white matter-segmented before being registered to the MNI152 standard space using non-linear registration. The resulting images were averaged and flipped along the X axis to create a left–right symmetric, study-specific grey matter template. To preserve spatial coherence between analyses, a white matter template was produced re-using registration parameters obtained for grey matter analysis. Second, all native grey and white matter images were nonlinearly registered to their respective study-specific template and ‘‘Jacobian-modulated’’ to correct for local expansion (or contraction) due to the spatial transformation. The modulated grey and white matter images were then smoothed with
J Neurol (2014) 261:291–299
an isotropic Gaussian kernel with a sigma of 4 mm. Finally, a voxel-wise general linear model (GLM) was applied using permutation testing, correcting for multiple comparisons across space. We set the alpha level at p = 0.05 for all tests in this study. Group differences were analysed with a two-sample T test using permutation testing, including age as a nuisance regressor to account for normal aging effects.
293
repeat length and (b) FA and MD values, respectively. Apart from age, this design included disease duration as a further nuisance regressor to remove effects not directly related to the measures under study. After regression analysis, the mean FA from voxels exhibiting significant correlation to SARA was calculated for each subject. We performed linear regression to compute the coefficient of determination (R2) between FA and SARA with SPSS 19 (IBM, Boulder, CO).
Data processing and analysis: diffusion Diffusion: diffusion tractography Data were corrected for shim coil heating effects, eddy currents and head motion by affine registration to a nondiffusion reference volume [21]. The three acquisitions were then averaged to increase SNR. Diffusion tensors were calculated with FSL, giving maps of FA and MD in the process. Furthermore, we extracted radial diffusivity (RD) and axial diffusivity (AD) as separate components of MD. To enable tractography, we calculated probability distributions of fibre directions for each brain voxel using a two-fibre model with FMRIB’s Diffusion Toolbox (FDT). We performed TBSS, including age as a nuisance regressor in all designs. TBSS projects all subjects’ FA, MD, RD and AD data onto an FA tract skeleton, before applying voxel-wise cross-subject statistics using permutation testing. We adjusted p values for multiple comparisons using threshold-free cluster enhancement [22]. To investigate group differences between the SCA3 and control set, we used a two-sample T test. Diffusion: regression analysis In the SCA3 group only, we performed voxel-wise multiple regression to test for correlations between (a) SARA score,
Using the voxels exhibiting a significant correlation between SARA and FA identified in the previous analysis, we then selected only those that lay within thalamus for further study, seeking to identify the portion of thalamus where this significant correlation was found. To this end, we intersected the parametric map with a standard mask of thalamus from the Harvard Oxford subcortical atlas, thus excluding all voxels outside thalamus from this analysis. The thalamic voxels identified were then back-projected into individual space for diffusion tractography. We performed probabilistic multi-fibre diffusion tractography from every voxel in this mask to identify remote targets of connectivity with the thalamic voxels whose FA correlates to SARA scores, releasing 5,000 virtual particles from each voxel. Counters were increased every time a virtual particle traversed any brain voxel, resulting in probabilistic maps of remote connectivity. A 95th percentile threshold was applied to these probabilistic connectivity maps to retain only voxels showing a high likelihood of connectivity to the thalamic seed voxels; the maps were then binarised and transformed into MNI152 standard space. We then summed up these maps to obtain population probability maps
Fig. 1 Grey matter (a) and white matter (b) loss in SCA3, as compared to normal controls, depicted in red-yellow. Note involvement of the cerebellum, pons, midbrain, and thalamus
123
294
of connectivity to the thalamic seed voxels, such that the value at each voxel would correspond to the number of subjects exhibiting evidence of connectivity to them.
Results In SCA3 patients, grey matter VBM detected loss of grey matter in the cerebellum, the brainstem including pons and midbrain, as well as bilaterally in the thalamus (Fig. 1a). White matter VBM demonstrated significant white matter loss in cerebellar, pontine and midbrain white matter, in the anterior midbrain and in thalamus bilaterally (Fig. 1b). Both grey and white matter VBM results were largely symmetric and closely related in space. Cluster characteristics are reported in supplementary Table 1. On between-group analysis of diffusion data between the SCA3 group and normal controls, TBSS identified widespread FA reduction in bilateral cerebral frontal, parietal, temporal, occipital and cerebellar white matter in SCA3 patients, also including thalamus and the brainstem (Fig. 2a). Moreover, MD increases were detected in a similar, widely overlapping pattern in bilateral frontal, parietal, temporal, thalamic and cerebellar white matter, and in the brainstem (Fig. 2b). Analysis of RD (Fig. 2c) and AD (Fig. 2d) identified increased RD as the main driving force behind the MD increase observed. Correlation analysis of the SARA motor score and diffusion-derived parameters detected a negative correlation between FA and SARA in bilateral frontal and thalamic white matter, and in the white matter of the midbrain and cerebellum (Fig. 2e). The coefficient of determination between the average FA in these regions across individual SCA3 patients and their SARA score was R2 = 0.615 (Fig. 3). Unlike the situation with FA, we did not find a correlation between MD and SARA. Furthermore, there was no correlation between trinucleotide repeat length and any of the imaging parameters. To identify the anatomical origin of the thalamic voxels exhibiting significant correlation between FA and SARA scores, these voxels were submitted to probabilistic diffusion tractography. We generated population probability maps depicting their remote connectivity. These maps showed robust connectivity between the thalamic seed voxels and the cerebellum, bilateral central sensorimotor regions, the basal ganglia, and frontal and mid-temporal brain regions (Fig. 4).
Discussion In this study, we employed non-invasive anatomical MRI and DWI to explore grey and white matter pathology in SCA3 patients.
123
J Neurol (2014) 261:291–299 Fig. 2 Between-group analysis of white matter diffusion imaging c detects significant decreases in white matter FA (a, colour-coded in red), and increases in MD (b, blue), AD (c, copper) and RD (d, yellow) in SCA3. Significant changes in FA (decrease), and MD, AD and RD (increase) are superimposed on the TBSS skeleton displayed in green. FA is correlated to SARA in a network of supra- and infratentorial white matter pathways including thalamic and frontal white matter (e). White matter regions exhibiting significant correlation between FA and SARA are superimposed in red on the TBSS skeleton plotted in green
In line with previously published imaging results [7, 23], we detected loss of grey matter in the cerebellum and brainstem including the pons and the midbrain. On inspection, cerebellar pathology involves the cerebellar cortex, and also the deep nuclei within cerebellum. The region containing the dentate nucleus, posterior to the middle cerebellar peduncle, is clearly involved in volume loss detected by VBM, mirroring findings of histopathology. However, the probabilistic nature of grey/white segmentation is a confounding factor in this region, as demonstrated by the apparent loss of ‘‘grey matter’’ in the middle cerebellar peduncle, and carry-over effects from the very prominent white matter loss probably contribute to the apparent grey matter loss found here. In supratentorial regions, we found reduced grey matter in thalamus, particularly in its ventrolateral aspects. This is in line with previous histological studies that have detected neuronal degeneration occurring in the ventrolateral group of thalamic nuclei in SCA3 [24, 25]. For white matter, we found volume loss in cerebellar and brainstem white matter, in line with previous reports [6, 11]. However, in addition to what has been reported previously, we also detected evidence of white matter loss in lateral thalamus, possibly a consequence of loss of neurons in the ventrolateral thalamus [24, 25] and subsequent axonal degeneration. Moreover, cerebellar projections terminate in this part of thalamus [26], and cerebellar degeneration is likely to result in disruption of these ascending axons. In addition to the atrophy just discussed, there was evidence of widespread microstructural damage to supra- and infratentorial white matter in SCA3 on between-group analysis of DWI. Remarkably, we detected widespread supratentorial FA decreases over and above the infratentorial FA decrease reported in a previous study [11]. This increase in sensitivity is likely due to the greater number of diffusion measurements taken here, making estimation of the diffusion tensor properties more robust [27]. Increases in MD and decreases in FA were largely collocalised (Fig. 2a, b), hinting at a common underlying process causing these changes of the diffusion profile DWI is sensitive to the diffusion of free water in brain tissue, providing a window into the microstructural properties of white matter in the living human brain. Diffusion
J Neurol (2014) 261:291–299
295
123
296
J Neurol (2014) 261:291–299
MRI can detect changes in white matter that do not necessarily result in gross macroscopic atrophy [12, 14]. Unfortunately, it is not straightforward to relate changes in diffusion characteristics to specific neuropathological changes, since the white matter’s diffusion signal depends on many factors, including axonal packing density, cellular
R 2 = 0.615
0.525
0.500
FA
0.475
0.450
0.425
0.400
0
5
10
15
20
SARA score
Fig. 3 Regression plot of each SCA3 patient’s individual average FA obtained from FA/SARA correlation analysis (cf. Fig. 2e), plotted against individual SARA scores, depicting a strong negative correlation between FA and SARA
Fig. 4 Population probability maps of connectivity with thalamic seed voxels found to exhibit correlation between FA and SARA (seeds depicted in inset). The colour bar indicates the number of subjects in our study group exhibiting seed connectivity in each brain voxel
123
membranes, intracellular organelles and myelin content of the tissue [28, 29]. However, reduction of FA, as opposed to a pure increase of MD, has been linked to axonal loss as an underlying mechanism [29]. As RD is the main component driving MD increase in this study (cf. Fig. 2d, e), we can assume that loss of cellular diffusion barriers, hindering diffusion perpendicular to the main fibre direction, occurs in our patient group. However, recent research suggests that interpretation of RD and AD changes alone can be ambiguous [30], and we need to take previous histopathological studies into consideration to interpret our findings. In SCA3, neurons in the cerebellum, brainstem and thalamus degenerate, where a large number of neural projections arise. In detail, neuropathological findings include neuronal loss and atrophy in the brainstem, the spinal cord and the cerebellar hemispheres including the dentate nucleus [24, 31–33]. Intranuclear inclusions or neuronal loss have also been observed in a number of thalamic nuclei [24, 25, 32, 34], the basal ganglia [35, 36] and the cerebral cortex [37]. It follows that the axons of these neurons will degenerate consecutively, providing one putative mechanism of the white matter involvement detected here. When interpreting our findings, it is important to remember that diffusion MRI cannot provide the level of evidence that histology methods will, and it cannot distinguish between pure Wallerian degeneration and other
J Neurol (2014) 261:291–299
modes of white matter dysintegrity. Other possible mechanisms are modes of ataxin-3 toxicity, be it through reduced degradation in the ubiquitin–proteasome system, through its interaction with transcription in the neuronal nucleus, or through a combination of these and other factors [38] leading to disruption of white matter integrity. Whatever the underlying pathogenetic mechanism, ataxin3’s ubiquitous presence in the brain makes it a prime candidate for the widespread changes observed here. Furthermore, the strong negative correlation observed between FA of a wide-spread supra- and infratentorial network and SARA ataxia scores in the SCA3 collective (Figs. 2e, 3) is clear indication that the white matter changes found in our study are functionally meaningful. The lateralisation in the cerebellum is probably spurious, but could alternatively be related to training-induced preservation of motor function in the usually dominant right hand, influencing the SARA score recorded. FA in the network described above provides an indicator of overall severity of motor symptoms. This fits well with the relationship previously found between loss of brain tissue and disease severity in a previous study using semiautomatic outlining of infratentorial brain structures [9]. The correlation we observed could thus either be due to degeneration of white matter consecutive to the loss of neurons in SCA3, but it is also possible that white matter pathology in itself contributes to the clinical presentation. In this context, it is intriguing to see that the white matter pathways in thalamus whose FA correlates to SARA scores support connectivity to the cerebellum and primary as well as premotor structures (Fig. 4). This particular pattern of connectivity has been shown to identify the posterior portion of the ventral lateral nucleus of thalamus (VLp) in previous studies [39, 40], an important relay in motor circuits [41]. Together with the grey matter reductions observed in the same region and the histopathological evidence discussed above, these findings suggest that both grey and white matter pathology of the ventrolateral nuclear group of thalamus is detectable with non-invasive MRI, and functionally relevant in SCA3.
Conclusion In conclusion, the core pathology of SCA3 leaves its mark on cerebral grey and white matter structures alike, and our data indicate that white matter involvement appears to be more extensive than previously thought, especially with respect to supratentorial structures. The severity of microstructural white matter damage in SCA3, as measured by FA, strongly correlates with the SARA ataxia severity scores in a specific cortico-subcortical network, including ventrolateral thalamus. Moreover, MRI is sensitive to
297
thalamic pathology on both a macroscopic and microscopic scale. Our data could imply a greater functional role for white matter damage, and might be useful as a surrogate parameter to assess disease status in patients with SCA3 in the future. Moreover, longitudinal studies will be needed to assess how well this method can track disease progression over time. Acknowledgments Frankfurt.
This study was funded by Goethe-University of
Conflicts of interest Jun-Suk Kang received honoraria and travel funding from GlaxoSmithKline, Ipsen Pharma, Merz Pharma, Teva Pharma, and Medtronic. Johannes C Klein reports no financial disclosures. Simon Baudrexel reports no financial disclosures. Dagmar Nolte reports no financial disclosures. Ruediger Hilker has received speaker honoraria from Medtronic, Orion, GlaxoSmithKline, TEVA, Cephalon, Solvay, Desitin, and Boehringer Ingelheim as well as travel funding from Medtronic and Cephalon; serves or has served on a scientific advisory board for Cephalon; and has received research funding from the Deutsche Parkinson Vereinigung (dPV), Bundesministerium fu¨r Bildung und Forschung and the Goethe-University of Frankfurt.
References 1. Rub U, Brunt ER, Deller T (2008) New insights into the pathoanatomy of spinocerebellar ataxia type 3 (Machado-Joseph disease). Curr Opin Neurol 21(2):111–116. doi:10.1097/WCO. 0b013e3282f7673d 2. Gatchel JR, Zoghbi HY (2005) Diseases of unstable repeat expansion: mechanisms and common principles. Nat Rev Genet 6(10):743–755. doi:10.1038/nrg1691 3. Paulson HL, Das SS, Crino PB, Perez MK, Patel SC, Gotsdiner D, Fischbeck KH, Pittman RN (1997) Machado-Joseph disease gene product is a cytoplasmic protein widely expressed in brain. Ann Neurol 41(4):453–462. doi:10.1002/ana.410410408 4. Ross CA (1997) Intranuclear neuronal inclusions: a common pathogenic mechanism for glutamine-repeat neurodegenerative diseases? Neuron 19(6):1147–1150 5. Schols L, Bauer P, Schmidt T, Schulte T, Riess O (2004) Autosomal dominant cerebellar ataxias: clinical features, genetics, and pathogenesis. Lancet Neurol 3(5):291–304. doi:10.1016/S14744422(04)00737-9 6. Lukas C, Schols L, Bellenberg B, Rub U, Przuntek H, Schmid G, Koster O, Suchan B (2006) Dissociation of grey and white matter reduction in spinocerebellar ataxia type 3 and 6: a voxel-based morphometry study. Neurosci Lett 408(3):230–235. doi:10.1016/ j.neulet.2006.09.007 7. Klockgether T, Skalej M, Wedekind D, Luft AR, Welte D, Schulz JB, Abele M, Burk K, Laccone F, Brice A, Dichgans J (1998) Autosomal dominant cerebellar ataxia type I. MRI-based volumetry of posterior fossa structures and basal ganglia in spinocerebellar ataxia types 1, 2 and 3. Brain 121(Pt 9):1687–1693 8. Jacobi H, Hauser TK, Giunti P, Globas C, Bauer P, SchmitzHubsch T, Baliko L, Filla A, Mariotti C, Rakowicz M, Charles P, Ribai P, Szymanski S, Infante J, van de Warrenburg BP, Durr A, Timmann D, Boesch S, Fancellu R, Rola R, Depondt C, Schols L, Zdzienicka E, Kang JS, Ratzka S, Kremer B, Stephenson DA, Melegh B, Pandolfo M, Tezenas du Montcel S, Borkert J, Schulz JB, Klockgether T (2012) Spinocerebellar ataxia types 1, 2, 3 and
123
298
9.
10.
11.
12.
13.
14.
15.
16.
17. 18.
19.
20.
21.
22.
J Neurol (2014) 261:291–299 6: the clinical spectrum of ataxia and morphometric brainstem and cerebellar findings. Cerebellum 11(1):155–166. doi:10.1007/ s12311-011-0292-z Schulz JB, Borkert J, Wolf S, Schmitz-Hubsch T, Rakowicz M, Mariotti C, Schols L, Timmann D, van de Warrenburg B, Durr A, Pandolfo M, Kang JS, Mandly AG, Nagele T, Grisoli M, Boguslawska R, Bauer P, Klockgether T, Hauser TK (2010) Visualization, quantification and correlation of brain atrophy with clinical symptoms in spinocerebellar ataxia types 1, 3 and 6. Neuroimage 49(1):158–168. doi:10.1016/j.neuroimage.2009.07. 027 D’Abreu A, Franca MC Jr, Yasuda CL, Campos BA, LopesCendes I, Cendes F (2012) Neocortical atrophy in MachadoJoseph disease: a longitudinal neuroimaging study. J Neuroimaging 22(3):285–291. doi:10.1111/j.1552-6569.2011.00614.x Guimaraes RP, D’Abreu A, Yasuda CL, Franca MC Jr, Silva BH, Cappabianco FA, Bergo FP, Lopes-Cendes IT, Cendes F (2013) A multimodal evaluation of microstructural white matter damage in spinocerebellar ataxia type 3. Mov Disord. doi:10.1002/mds. 25451 Ciccarelli O, Behrens TE, Altmann DR, Orrell RW, Howard RS, Johansen-Berg H, Miller DH, Matthews PM, Thompson AJ (2006) Probabilistic diffusion tractography: a potential tool to assess the rate of disease progression in amyotrophic lateral sclerosis. Brain 129:1859–1871 Benedetti B, Rocca MA, Rovaris M, Caputo D, Zaffaroni M, Capra R, Bertolotto A, Martinelli V, Comi G, Filippi M (2010) A diffusion tensor MRI study of cervical cord damage in benign and secondary progressive multiple sclerosis patients. J Neurol Neurosurg Psychiatry 81(1):26–30. doi:10.1136/jnnp.2009.173120 Klein JC, Lorenz B, Kang JS, Baudrexel S, Seifried C, van de Loo S, Steinmetz H, Deichmann R, Hilker R (2011) Diffusion tensor imaging of white matter involvement in essential tremor. Hum Brain Mapp 32(6):896–904. doi:10.1002/hbm.21077 Schmitz-Hubsch T, du Montcel ST, Baliko L, Berciano J, Boesch S, Depondt C, Giunti P, Globas C, Infante J, Kang JS, Kremer B, Mariotti C, Melegh B, Pandolfo M, Rakowicz M, Ribai P, Rola R, Schols L, Szymanski S, van de Warrenburg BP, Durr A, Klockgether T, Fancellu R (2006) Scale for the assessment and rating of ataxia: development of a new clinical scale. Neurology 66(11):1717–1720. doi:10.1212/01.wnl.0000219042.60538.92 Frahm J, Haase A, Matthaei D (1986) Rapid NMR imaging of dynamic processes using the FLASH technique. Magn Reson Med 3(2):321–327 Turner R, Le Bihan D, Chesnick AS (1991) Echo-planar imaging of diffusion and perfusion. Magn Reson Med 19(2):247–253 Griswold MA, Jakob PM, Heidemann RM, Nittka M, Jellus V, Wang J, Kiefer B, Haase A (2002) Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 47(6):1202–1210. doi:10.1002/mrm.10171 Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, Johansen-Berg H, Bannister PR, De Luca M, Drobnjak I, Flitney DE, Niazy RK, Saunders J, Vickers J, Zhang Y, De Stefano N, Brady JM, Matthews PM (2004) Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage 23(Suppl 1):S208–S219 Good CD, Johnsrude IS, Ashburner J, Henson RN, Friston KJ, Frackowiak RS (2001) A voxel-based morphometric study of ageing in 465 normal adult human brains. Neuroimage 14(1 Pt 1):21–36. doi:10.1006/nimg 2001.0786 Jenkinson M, Bannister P, Brady M, Smith S (2002) Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage 17(2):825–841 Smith SM, Jenkinson M, Johansen-Berg H, Rueckert D, Nichols TE, Mackay CE, Watkins KE, Ciccarelli O, Cader MZ, Matthews PM, Behrens TEJ (2006) Tract-based spatial statistics: voxelwise
123
23.
24.
25.
26.
27.
28. 29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
analysis of multi-subject diffusion data. Neuroimage 31(4):1487– 1505 Eichler L, Bellenberg B, Hahn HK, Koster O, Schols L, Lukas C (2011) Quantitative assessment of brain stem and cerebellar atrophy in spinocerebellar ataxia types 3 and 6: impact on clinical status. AJNR Am J Neuroradiol 32(5):890–897. doi:10.3174/ajnr. A2387 Rub U, Brunt ER, Petrasch-Parwez E, Schols L, Theegarten D, Auburger G, Seidel K, Schultz C, Gierga K, Paulson H, van Broeckhoven C, Deller T, de Vos RA (2006) Degeneration of ingestion-related brainstem nuclei in spinocerebellar ataxia type 2, 3, 6 and 7. Neuropathol Appl Neurobiol 32(6):635–649. doi:10. 1111/j.1365-2990.2006.00772.x Rub U, Del Turco D, Del Tredici K, de Vos RA, Brunt ER, Reifenberger G, Seifried C, Schultz C, Auburger G, Braak H (2003) Thalamic involvement in a spinocerebellar ataxia type 2 (SCA2) and a spinocerebellar ataxia type 3 (SCA3) patient, and its clinical relevance. Brain 126(Pt 10):2257–2272. doi:10.1093/ brain/awg234 Asanuma C, Thach WT, Jones EG (1983) Distribution of cerebellar terminations and their relation to other afferent terminations in the ventral lateral thalamic region of the monkey. Brain Res 286(3):237–265 Jones DK, Horsfield MA, Simmons A (1999) Optimal strategies for measuring diffusion in anisotropic systems by magnetic resonance imaging. Magn Reson Med 42(3):515–525 Basser PJ, Mattiello J, Lebihan D (1994) MR diffusion tensor spectroscopy and imaging. Biophys J 66(1):259–267 Beaulieu C (2009) The biological basis of diffusion anisotropy. In: Johansen-Berg H, Behrens TEJ (eds) Diffusion MRI. Academic Press, Oxford, pp 105–126 Wheeler-Kingshott CA, Cercignani M (2009) About ‘‘axial’’ and ‘‘radial’’ diffusivities. Magn Reson Med 61(5):1255–1260. doi:10.1002/mrm.21965 Rub U, Gierga K, Brunt ER, de Vos RA, Bauer M, Schols L, Burk K, Auburger G, Bohl J, Schultz C, Vuksic M, Burbach GJ, Braak H, Deller T (2005) Spinocerebellar ataxias types 2 and 3: degeneration of the pre-cerebellar nuclei isolates the three phylogenetically defined regions of the cerebellum. J Neural Transm 112(11):1523–1545. doi:10.1007/s00702-005-0287-3 Yamada M, Tan CF, Inenaga C, Tsuji S, Takahashi H (2004) Sharing of polyglutamine localization by the neuronal nucleus and cytoplasm in CAG-repeat diseases. Neuropathol Appl Neurobiol 30(6):665–675. doi:10.1111/j.1365-2990.2004.00583.x Iwabuchi K, Tsuchiya K, Uchihara T, Yagishita S (1999) Autosomal dominant spinocerebellar degenerations. Clinical, pathological, and genetic correlations. Rev Neurol 155(4):255–270 Evert BO, Schelhaas J, Fleischer H, de Vos RA, Brunt ER, Stenzel W, Klockgether T, Wullner U (2006) Neuronal intranuclear inclusions, dysregulation of cytokine expression and cell death in spinocerebellar ataxia type 3. Clin Neuropathol 25(6):272–281 Durr A, Stevanin G, Cancel G, Duyckaerts C, Abbas N, Didierjean O, Chneiweiss H, Benomar A, Lyon-Caen O, Julien J, Serdaru M, Penet C, Agid Y, Brice A (1996) Spinocerebellar ataxia 3 and Machado-Joseph disease: clinical, molecular, and neuropathological features. Ann Neurol 39(4):490–499. doi:10.1002/ ana.410390411 Yamada M, Sato T, Tsuji S, Takahashi H (2008) CAG repeat disorder models and human neuropathology: similarities and differences. Acta Neuropathol 115(1):71–86. doi:10.1007/ s00401-007-0287-5 Yamada M, Hayashi S, Tsuji S, Takahashi H (2001) Involvement of the cerebral cortex and autonomic ganglia in Machado-Joseph disease. Acta Neuropathol 101(2):140–144 Matos CA, de Macedo-Ribeiro S, Carvalho AL (2011) Polyglutamine diseases: the special case of ataxin-3 and Machado-Joseph
J Neurol (2014) 261:291–299 disease. Prog Neurobiol 95(1):26–48. doi:10.1016/j.pneurobio. 2011.06.007 39. Johansen-Berg H, Behrens TEJ, Sillery E, Ciccarelli O, Thompson AJ, Smith SM, Matthews PM (2005) Functional-anatomical validation and individual variation of diffusion tractographybased segmentation of the human thalamus. Cereb Cortex 15(1):31–39 40. Klein JC, Barbe MT, Seifried C, Baudrexel S, Runge M, Maarouf M, Gasser T, Hattingen E, Liebig T, Deichmann R, Timmermann
299 L, Weise L, Hilker R (2012) The tremor network targeted by successful VIM deep brain stimulation in humans. Neurology 78(11):787–795. doi:10.1212/WNL.0b013e318249f702 41. Percheron G, Francois C, Talbi B, Meder JF, Fenelon G, Yelnik J (1993) The primate motor thalamus analysed with reference to subcortical afferent territories. Stereotact Funct Neurosurg 60(1– 3):32–41
123
ORIGINAL COMMUNICATION
White matter damage is related to ataxia severity in SCA3 J.-S. Kang • J. C. Klein • S. Baudrexel R. Deichmann • D. Nolte • R. Hilker
•
Received: 5 September 2013 / Revised: 5 November 2013 / Accepted: 5 November 2013 / Published online: 23 November 2013 Ó Springer-Verlag Berlin Heidelberg 2013
Abstract Spinocerebellar ataxia type 3 (SCA3) is the most frequent inherited cerebellar ataxia in Europe, the US and Japan, leading to disability and death through motor complications. Although the affected protein ataxin-3 is found ubiquitously in the brain, grey matter atrophy is predominant in the cerebellum and the brainstem. White matter pathology is generally less severe and thought to occur in the brainstem, spinal cord, and cerebellar white matter. Here, we investigated both grey and white matter pathology in a group of 12 SCA3 patients and matched controls. We used voxel-based morphometry for analysis of tissue loss, and tract-based spatial statistics (TBSS) on diffusion magnetic resonance imaging to investigate microstructural pathology. We analysed correlations between microstructural properties of the brain and ataxia severity, as measured by the Scale for the Assessment and Rating of Ataxia (SARA) score. SCA3 patients exhibited significant loss of both grey and white matter in the cerebellar hemispheres, brainstem including pons and in lateral
thalamus. On between-group analysis, TBSS detected widespread microstructural white matter pathology in the cerebellum, brainstem, and bilaterally in thalamus and the cerebral hemispheres. Furthermore, fractional anisotropy in a white matter network comprising frontal, thalamic, brainstem and left cerebellar white matter strongly and negatively correlated with SARA ataxia scores. Tractography identified the thalamic white matter thus implicated as belonging to ventrolateral thalamus. Disruption of white matter integrity in patients suffering from SCA3 is more widespread than previously thought. Moreover, our data provide evidence that microstructural white matter changes in SCA3 are strongly related to the clinical severity of ataxia symptoms. Keywords Spinocerebellar ataxia 3 MachadoJoseph disease DTI Diffusion Ataxia
Introduction J.S. Kang and J.C. Klein equally contributed to this work.
Electronic supplementary material The online version of this article (doi:10.1007/s00415-013-7186-6) contains supplementary material, which is available to authorized users. J.-S. Kang J. C. Klein S. Baudrexel R. Hilker Department of Neurology, Goethe-University of Frankfurt, Schleusenweg 2-16, 60528 Frankfurt am Main, Germany J.-S. Kang J. C. Klein (&) S. Baudrexel R. Deichmann Brain Imaging Center (BIC), Goethe-University of Frankfurt, Schleusenweg 2-16, 60528 Frankfurt am Main, Germany e-mail: [email protected] D. Nolte Institute of Human Genetics, University of Giessen, Schlangenzahl 14, 35392 Giessen, Germany
Spinocerebellar ataxia type 3 (SCA3), also known as Machado-Joseph disease, is the most frequent inherited cerebellar ataxia in Europe, the US and Japan [1]. SCA3 is an autosomal-dominant neurodegenerative disease with a variable clinical presentation, comprising ataxia, spasticity, parkinsonism, neuropathy, dystonia, and ophthalmoplegia [1]. SCA3 is caused by an abnormal CAG trinucleotide expansion on chromosome 14 (gene locus 14q32.12), encoding for polyglutamine in the abnormal gene product ataxin-3. The repeat length of the pathologically expanded allele varies between 45 and 51 repeats with reduced penetrance, and 52 and 86 copies with full penetrance. However, the exact pathogenesis of SCA3 is currently
123
292
unknown. A gain of function of the ‘‘toxic protein’’ is one possible pathomechanism discussed [2]. Neuronal intranuclear inclusion bodies are a pathological hallmark at autopsy, and may also play a role in the pathogenesis of SCA3 [3, 4]. A greater repeat length seems to precipitate an earlier disease onset [5]. Although ataxin-3 is found ubiquitously in the brain, macroscopic brain atrophy has been shown predominantly in the cerebellum and in the brainstem [6]. Neuroimaging studies, mostly employing standard magnetic resonance imaging (MRI), revealed marked cerebellar and pontine atrophy in SCA3 patients [7, 8]. The amount of brain atrophy correlated with the disease severity, and weakly with CAG repeat length [9]. Recently, evidence for cortical involvement has been reported in a longitudinal imaging study measuring cortical thickness [10]. In contrast, white matter pathology was generally considered less severe and, on histology, thought to be largely limited to brainstem, spinal cord, and cerebellar white matter [6]. However, one recent diffusion-weighted MRI (DWI) study demonstrated increased water diffusivity in supratentorial white matter [11]. In this observational study, we aimed to investigate both grey and white matter pathology in a group of 12 subjects with genetically proven SCA3. For this, we performed voxel-based morphometry on optimised anatomical MRI, and tract-based spatial statistics (TBSS) as well as tractography on DWI. Measures derived from the diffusion tensor model, such as fractional anisotropy (FA) and mean diffusivity (MD), have been shown to relate to the severity of degenerative and inflammatory diseases of the brain [12–14]. Therefore, we set out to explore white matter dysintegrity in SCA3 and, in particular, its relationship with clinically evident motor impairment.
Patients and methods Subjects Twelve patients with a molecular diagnosis of SCA3 participated in this study (five women, seven men; 50.5 ± 10.4 years, range 30–63 years). The mean CAG repeat length was 70.6 ± 3.8, range 65–78 copies. Mean disease duration was 11.0 ± 6.4 years and the mean Scale for the Assessment and Rating of Ataxia (SARA) [15] score amounted to 10.3 ± 4.6. Twelve age- and sex-matched control subjects (five women, seven men; 47.8 ± 11.02 years, ranging from 26 to 61 years) participated as a control group. None of the participants had a history of other neurological or psychiatric disease, and neurological examination was normal in the control group. All participants gave their informed consent before entering
123
J Neurol (2014) 261:291–299
the study. The study conformed to the latest revision of the declaration of Helsinki of 1995 and was approved by the ethics committee of the University Hospital Frankfurt. MR imaging All subjects underwent MRI on a Siemens Trio 3T (Siemens, Erlangen, Germany) scanner at Frankfurt University’s Brain Imaging Center. For all MR procedures, the head was immobilised using self-expanding foam cushions. Image acquisition used an 8-channel array head coil for signal reception, and the body coil for transmission. Anatomical images were recorded using T1-weighted spoiled FLASH (TE 2.4 ms, TR 7.6 ms, flip angle 18°, 176 sagittal slices, matrix 224 9 256, isotropic voxel size 1 mm, 2 repeats) [16]. Diffusion was measured along 60 isotropically distributed directions using spin-echo EPI [17] and GRAPPA [18] (SE-EPI, TE 95 ms, TR 9.3 s, 70 axial slices, matrix 104 9 104, isotropic voxel size 2 mm, acceleration factor 2, 33 reference lines, b value 1,000 s mm-2) and 10 non-diffusion weighted reference images. Diffusion scanning was repeated three times to increase SNR. Common space All data processing used tools from the FMRIB software library (FSL) [19]. To enable between-subject statistics, all images were non-linearly registered to a high-resolution MR template in MNI152 space provided with FSL. Parameters taking individual images into standard space and vice versa were recorded to enable analysis across individual imaging spaces. Data processing and analysis: VBM grey and white matter First, the two FLASH volumes were registered linearly to correct for potential subject movement. These volumes were then summed to reduce noise while maintaining high grey to white matter contrast. Anatomical data were then analysed with FSL-VBM [20]. First, images were brain-extracted and grey and white matter-segmented before being registered to the MNI152 standard space using non-linear registration. The resulting images were averaged and flipped along the X axis to create a left–right symmetric, study-specific grey matter template. To preserve spatial coherence between analyses, a white matter template was produced re-using registration parameters obtained for grey matter analysis. Second, all native grey and white matter images were nonlinearly registered to their respective study-specific template and ‘‘Jacobian-modulated’’ to correct for local expansion (or contraction) due to the spatial transformation. The modulated grey and white matter images were then smoothed with
J Neurol (2014) 261:291–299
an isotropic Gaussian kernel with a sigma of 4 mm. Finally, a voxel-wise general linear model (GLM) was applied using permutation testing, correcting for multiple comparisons across space. We set the alpha level at p = 0.05 for all tests in this study. Group differences were analysed with a two-sample T test using permutation testing, including age as a nuisance regressor to account for normal aging effects.
293
repeat length and (b) FA and MD values, respectively. Apart from age, this design included disease duration as a further nuisance regressor to remove effects not directly related to the measures under study. After regression analysis, the mean FA from voxels exhibiting significant correlation to SARA was calculated for each subject. We performed linear regression to compute the coefficient of determination (R2) between FA and SARA with SPSS 19 (IBM, Boulder, CO).
Data processing and analysis: diffusion Diffusion: diffusion tractography Data were corrected for shim coil heating effects, eddy currents and head motion by affine registration to a nondiffusion reference volume [21]. The three acquisitions were then averaged to increase SNR. Diffusion tensors were calculated with FSL, giving maps of FA and MD in the process. Furthermore, we extracted radial diffusivity (RD) and axial diffusivity (AD) as separate components of MD. To enable tractography, we calculated probability distributions of fibre directions for each brain voxel using a two-fibre model with FMRIB’s Diffusion Toolbox (FDT). We performed TBSS, including age as a nuisance regressor in all designs. TBSS projects all subjects’ FA, MD, RD and AD data onto an FA tract skeleton, before applying voxel-wise cross-subject statistics using permutation testing. We adjusted p values for multiple comparisons using threshold-free cluster enhancement [22]. To investigate group differences between the SCA3 and control set, we used a two-sample T test. Diffusion: regression analysis In the SCA3 group only, we performed voxel-wise multiple regression to test for correlations between (a) SARA score,
Using the voxels exhibiting a significant correlation between SARA and FA identified in the previous analysis, we then selected only those that lay within thalamus for further study, seeking to identify the portion of thalamus where this significant correlation was found. To this end, we intersected the parametric map with a standard mask of thalamus from the Harvard Oxford subcortical atlas, thus excluding all voxels outside thalamus from this analysis. The thalamic voxels identified were then back-projected into individual space for diffusion tractography. We performed probabilistic multi-fibre diffusion tractography from every voxel in this mask to identify remote targets of connectivity with the thalamic voxels whose FA correlates to SARA scores, releasing 5,000 virtual particles from each voxel. Counters were increased every time a virtual particle traversed any brain voxel, resulting in probabilistic maps of remote connectivity. A 95th percentile threshold was applied to these probabilistic connectivity maps to retain only voxels showing a high likelihood of connectivity to the thalamic seed voxels; the maps were then binarised and transformed into MNI152 standard space. We then summed up these maps to obtain population probability maps
Fig. 1 Grey matter (a) and white matter (b) loss in SCA3, as compared to normal controls, depicted in red-yellow. Note involvement of the cerebellum, pons, midbrain, and thalamus
123
294
of connectivity to the thalamic seed voxels, such that the value at each voxel would correspond to the number of subjects exhibiting evidence of connectivity to them.
Results In SCA3 patients, grey matter VBM detected loss of grey matter in the cerebellum, the brainstem including pons and midbrain, as well as bilaterally in the thalamus (Fig. 1a). White matter VBM demonstrated significant white matter loss in cerebellar, pontine and midbrain white matter, in the anterior midbrain and in thalamus bilaterally (Fig. 1b). Both grey and white matter VBM results were largely symmetric and closely related in space. Cluster characteristics are reported in supplementary Table 1. On between-group analysis of diffusion data between the SCA3 group and normal controls, TBSS identified widespread FA reduction in bilateral cerebral frontal, parietal, temporal, occipital and cerebellar white matter in SCA3 patients, also including thalamus and the brainstem (Fig. 2a). Moreover, MD increases were detected in a similar, widely overlapping pattern in bilateral frontal, parietal, temporal, thalamic and cerebellar white matter, and in the brainstem (Fig. 2b). Analysis of RD (Fig. 2c) and AD (Fig. 2d) identified increased RD as the main driving force behind the MD increase observed. Correlation analysis of the SARA motor score and diffusion-derived parameters detected a negative correlation between FA and SARA in bilateral frontal and thalamic white matter, and in the white matter of the midbrain and cerebellum (Fig. 2e). The coefficient of determination between the average FA in these regions across individual SCA3 patients and their SARA score was R2 = 0.615 (Fig. 3). Unlike the situation with FA, we did not find a correlation between MD and SARA. Furthermore, there was no correlation between trinucleotide repeat length and any of the imaging parameters. To identify the anatomical origin of the thalamic voxels exhibiting significant correlation between FA and SARA scores, these voxels were submitted to probabilistic diffusion tractography. We generated population probability maps depicting their remote connectivity. These maps showed robust connectivity between the thalamic seed voxels and the cerebellum, bilateral central sensorimotor regions, the basal ganglia, and frontal and mid-temporal brain regions (Fig. 4).
Discussion In this study, we employed non-invasive anatomical MRI and DWI to explore grey and white matter pathology in SCA3 patients.
123
J Neurol (2014) 261:291–299 Fig. 2 Between-group analysis of white matter diffusion imaging c detects significant decreases in white matter FA (a, colour-coded in red), and increases in MD (b, blue), AD (c, copper) and RD (d, yellow) in SCA3. Significant changes in FA (decrease), and MD, AD and RD (increase) are superimposed on the TBSS skeleton displayed in green. FA is correlated to SARA in a network of supra- and infratentorial white matter pathways including thalamic and frontal white matter (e). White matter regions exhibiting significant correlation between FA and SARA are superimposed in red on the TBSS skeleton plotted in green
In line with previously published imaging results [7, 23], we detected loss of grey matter in the cerebellum and brainstem including the pons and the midbrain. On inspection, cerebellar pathology involves the cerebellar cortex, and also the deep nuclei within cerebellum. The region containing the dentate nucleus, posterior to the middle cerebellar peduncle, is clearly involved in volume loss detected by VBM, mirroring findings of histopathology. However, the probabilistic nature of grey/white segmentation is a confounding factor in this region, as demonstrated by the apparent loss of ‘‘grey matter’’ in the middle cerebellar peduncle, and carry-over effects from the very prominent white matter loss probably contribute to the apparent grey matter loss found here. In supratentorial regions, we found reduced grey matter in thalamus, particularly in its ventrolateral aspects. This is in line with previous histological studies that have detected neuronal degeneration occurring in the ventrolateral group of thalamic nuclei in SCA3 [24, 25]. For white matter, we found volume loss in cerebellar and brainstem white matter, in line with previous reports [6, 11]. However, in addition to what has been reported previously, we also detected evidence of white matter loss in lateral thalamus, possibly a consequence of loss of neurons in the ventrolateral thalamus [24, 25] and subsequent axonal degeneration. Moreover, cerebellar projections terminate in this part of thalamus [26], and cerebellar degeneration is likely to result in disruption of these ascending axons. In addition to the atrophy just discussed, there was evidence of widespread microstructural damage to supra- and infratentorial white matter in SCA3 on between-group analysis of DWI. Remarkably, we detected widespread supratentorial FA decreases over and above the infratentorial FA decrease reported in a previous study [11]. This increase in sensitivity is likely due to the greater number of diffusion measurements taken here, making estimation of the diffusion tensor properties more robust [27]. Increases in MD and decreases in FA were largely collocalised (Fig. 2a, b), hinting at a common underlying process causing these changes of the diffusion profile DWI is sensitive to the diffusion of free water in brain tissue, providing a window into the microstructural properties of white matter in the living human brain. Diffusion
J Neurol (2014) 261:291–299
295
123
296
J Neurol (2014) 261:291–299
MRI can detect changes in white matter that do not necessarily result in gross macroscopic atrophy [12, 14]. Unfortunately, it is not straightforward to relate changes in diffusion characteristics to specific neuropathological changes, since the white matter’s diffusion signal depends on many factors, including axonal packing density, cellular
R 2 = 0.615
0.525
0.500
FA
0.475
0.450
0.425
0.400
0
5
10
15
20
SARA score
Fig. 3 Regression plot of each SCA3 patient’s individual average FA obtained from FA/SARA correlation analysis (cf. Fig. 2e), plotted against individual SARA scores, depicting a strong negative correlation between FA and SARA
Fig. 4 Population probability maps of connectivity with thalamic seed voxels found to exhibit correlation between FA and SARA (seeds depicted in inset). The colour bar indicates the number of subjects in our study group exhibiting seed connectivity in each brain voxel
123
membranes, intracellular organelles and myelin content of the tissue [28, 29]. However, reduction of FA, as opposed to a pure increase of MD, has been linked to axonal loss as an underlying mechanism [29]. As RD is the main component driving MD increase in this study (cf. Fig. 2d, e), we can assume that loss of cellular diffusion barriers, hindering diffusion perpendicular to the main fibre direction, occurs in our patient group. However, recent research suggests that interpretation of RD and AD changes alone can be ambiguous [30], and we need to take previous histopathological studies into consideration to interpret our findings. In SCA3, neurons in the cerebellum, brainstem and thalamus degenerate, where a large number of neural projections arise. In detail, neuropathological findings include neuronal loss and atrophy in the brainstem, the spinal cord and the cerebellar hemispheres including the dentate nucleus [24, 31–33]. Intranuclear inclusions or neuronal loss have also been observed in a number of thalamic nuclei [24, 25, 32, 34], the basal ganglia [35, 36] and the cerebral cortex [37]. It follows that the axons of these neurons will degenerate consecutively, providing one putative mechanism of the white matter involvement detected here. When interpreting our findings, it is important to remember that diffusion MRI cannot provide the level of evidence that histology methods will, and it cannot distinguish between pure Wallerian degeneration and other
J Neurol (2014) 261:291–299
modes of white matter dysintegrity. Other possible mechanisms are modes of ataxin-3 toxicity, be it through reduced degradation in the ubiquitin–proteasome system, through its interaction with transcription in the neuronal nucleus, or through a combination of these and other factors [38] leading to disruption of white matter integrity. Whatever the underlying pathogenetic mechanism, ataxin3’s ubiquitous presence in the brain makes it a prime candidate for the widespread changes observed here. Furthermore, the strong negative correlation observed between FA of a wide-spread supra- and infratentorial network and SARA ataxia scores in the SCA3 collective (Figs. 2e, 3) is clear indication that the white matter changes found in our study are functionally meaningful. The lateralisation in the cerebellum is probably spurious, but could alternatively be related to training-induced preservation of motor function in the usually dominant right hand, influencing the SARA score recorded. FA in the network described above provides an indicator of overall severity of motor symptoms. This fits well with the relationship previously found between loss of brain tissue and disease severity in a previous study using semiautomatic outlining of infratentorial brain structures [9]. The correlation we observed could thus either be due to degeneration of white matter consecutive to the loss of neurons in SCA3, but it is also possible that white matter pathology in itself contributes to the clinical presentation. In this context, it is intriguing to see that the white matter pathways in thalamus whose FA correlates to SARA scores support connectivity to the cerebellum and primary as well as premotor structures (Fig. 4). This particular pattern of connectivity has been shown to identify the posterior portion of the ventral lateral nucleus of thalamus (VLp) in previous studies [39, 40], an important relay in motor circuits [41]. Together with the grey matter reductions observed in the same region and the histopathological evidence discussed above, these findings suggest that both grey and white matter pathology of the ventrolateral nuclear group of thalamus is detectable with non-invasive MRI, and functionally relevant in SCA3.
Conclusion In conclusion, the core pathology of SCA3 leaves its mark on cerebral grey and white matter structures alike, and our data indicate that white matter involvement appears to be more extensive than previously thought, especially with respect to supratentorial structures. The severity of microstructural white matter damage in SCA3, as measured by FA, strongly correlates with the SARA ataxia severity scores in a specific cortico-subcortical network, including ventrolateral thalamus. Moreover, MRI is sensitive to
297
thalamic pathology on both a macroscopic and microscopic scale. Our data could imply a greater functional role for white matter damage, and might be useful as a surrogate parameter to assess disease status in patients with SCA3 in the future. Moreover, longitudinal studies will be needed to assess how well this method can track disease progression over time. Acknowledgments Frankfurt.
This study was funded by Goethe-University of
Conflicts of interest Jun-Suk Kang received honoraria and travel funding from GlaxoSmithKline, Ipsen Pharma, Merz Pharma, Teva Pharma, and Medtronic. Johannes C Klein reports no financial disclosures. Simon Baudrexel reports no financial disclosures. Dagmar Nolte reports no financial disclosures. Ruediger Hilker has received speaker honoraria from Medtronic, Orion, GlaxoSmithKline, TEVA, Cephalon, Solvay, Desitin, and Boehringer Ingelheim as well as travel funding from Medtronic and Cephalon; serves or has served on a scientific advisory board for Cephalon; and has received research funding from the Deutsche Parkinson Vereinigung (dPV), Bundesministerium fu¨r Bildung und Forschung and the Goethe-University of Frankfurt.
References 1. Rub U, Brunt ER, Deller T (2008) New insights into the pathoanatomy of spinocerebellar ataxia type 3 (Machado-Joseph disease). Curr Opin Neurol 21(2):111–116. doi:10.1097/WCO. 0b013e3282f7673d 2. Gatchel JR, Zoghbi HY (2005) Diseases of unstable repeat expansion: mechanisms and common principles. Nat Rev Genet 6(10):743–755. doi:10.1038/nrg1691 3. Paulson HL, Das SS, Crino PB, Perez MK, Patel SC, Gotsdiner D, Fischbeck KH, Pittman RN (1997) Machado-Joseph disease gene product is a cytoplasmic protein widely expressed in brain. Ann Neurol 41(4):453–462. doi:10.1002/ana.410410408 4. Ross CA (1997) Intranuclear neuronal inclusions: a common pathogenic mechanism for glutamine-repeat neurodegenerative diseases? Neuron 19(6):1147–1150 5. Schols L, Bauer P, Schmidt T, Schulte T, Riess O (2004) Autosomal dominant cerebellar ataxias: clinical features, genetics, and pathogenesis. Lancet Neurol 3(5):291–304. doi:10.1016/S14744422(04)00737-9 6. Lukas C, Schols L, Bellenberg B, Rub U, Przuntek H, Schmid G, Koster O, Suchan B (2006) Dissociation of grey and white matter reduction in spinocerebellar ataxia type 3 and 6: a voxel-based morphometry study. Neurosci Lett 408(3):230–235. doi:10.1016/ j.neulet.2006.09.007 7. Klockgether T, Skalej M, Wedekind D, Luft AR, Welte D, Schulz JB, Abele M, Burk K, Laccone F, Brice A, Dichgans J (1998) Autosomal dominant cerebellar ataxia type I. MRI-based volumetry of posterior fossa structures and basal ganglia in spinocerebellar ataxia types 1, 2 and 3. Brain 121(Pt 9):1687–1693 8. Jacobi H, Hauser TK, Giunti P, Globas C, Bauer P, SchmitzHubsch T, Baliko L, Filla A, Mariotti C, Rakowicz M, Charles P, Ribai P, Szymanski S, Infante J, van de Warrenburg BP, Durr A, Timmann D, Boesch S, Fancellu R, Rola R, Depondt C, Schols L, Zdzienicka E, Kang JS, Ratzka S, Kremer B, Stephenson DA, Melegh B, Pandolfo M, Tezenas du Montcel S, Borkert J, Schulz JB, Klockgether T (2012) Spinocerebellar ataxia types 1, 2, 3 and
123
298
9.
10.
11.
12.
13.
14.
15.
16.
17. 18.
19.
20.
21.
22.
J Neurol (2014) 261:291–299 6: the clinical spectrum of ataxia and morphometric brainstem and cerebellar findings. Cerebellum 11(1):155–166. doi:10.1007/ s12311-011-0292-z Schulz JB, Borkert J, Wolf S, Schmitz-Hubsch T, Rakowicz M, Mariotti C, Schols L, Timmann D, van de Warrenburg B, Durr A, Pandolfo M, Kang JS, Mandly AG, Nagele T, Grisoli M, Boguslawska R, Bauer P, Klockgether T, Hauser TK (2010) Visualization, quantification and correlation of brain atrophy with clinical symptoms in spinocerebellar ataxia types 1, 3 and 6. Neuroimage 49(1):158–168. doi:10.1016/j.neuroimage.2009.07. 027 D’Abreu A, Franca MC Jr, Yasuda CL, Campos BA, LopesCendes I, Cendes F (2012) Neocortical atrophy in MachadoJoseph disease: a longitudinal neuroimaging study. J Neuroimaging 22(3):285–291. doi:10.1111/j.1552-6569.2011.00614.x Guimaraes RP, D’Abreu A, Yasuda CL, Franca MC Jr, Silva BH, Cappabianco FA, Bergo FP, Lopes-Cendes IT, Cendes F (2013) A multimodal evaluation of microstructural white matter damage in spinocerebellar ataxia type 3. Mov Disord. doi:10.1002/mds. 25451 Ciccarelli O, Behrens TE, Altmann DR, Orrell RW, Howard RS, Johansen-Berg H, Miller DH, Matthews PM, Thompson AJ (2006) Probabilistic diffusion tractography: a potential tool to assess the rate of disease progression in amyotrophic lateral sclerosis. Brain 129:1859–1871 Benedetti B, Rocca MA, Rovaris M, Caputo D, Zaffaroni M, Capra R, Bertolotto A, Martinelli V, Comi G, Filippi M (2010) A diffusion tensor MRI study of cervical cord damage in benign and secondary progressive multiple sclerosis patients. J Neurol Neurosurg Psychiatry 81(1):26–30. doi:10.1136/jnnp.2009.173120 Klein JC, Lorenz B, Kang JS, Baudrexel S, Seifried C, van de Loo S, Steinmetz H, Deichmann R, Hilker R (2011) Diffusion tensor imaging of white matter involvement in essential tremor. Hum Brain Mapp 32(6):896–904. doi:10.1002/hbm.21077 Schmitz-Hubsch T, du Montcel ST, Baliko L, Berciano J, Boesch S, Depondt C, Giunti P, Globas C, Infante J, Kang JS, Kremer B, Mariotti C, Melegh B, Pandolfo M, Rakowicz M, Ribai P, Rola R, Schols L, Szymanski S, van de Warrenburg BP, Durr A, Klockgether T, Fancellu R (2006) Scale for the assessment and rating of ataxia: development of a new clinical scale. Neurology 66(11):1717–1720. doi:10.1212/01.wnl.0000219042.60538.92 Frahm J, Haase A, Matthaei D (1986) Rapid NMR imaging of dynamic processes using the FLASH technique. Magn Reson Med 3(2):321–327 Turner R, Le Bihan D, Chesnick AS (1991) Echo-planar imaging of diffusion and perfusion. Magn Reson Med 19(2):247–253 Griswold MA, Jakob PM, Heidemann RM, Nittka M, Jellus V, Wang J, Kiefer B, Haase A (2002) Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magn Reson Med 47(6):1202–1210. doi:10.1002/mrm.10171 Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, Johansen-Berg H, Bannister PR, De Luca M, Drobnjak I, Flitney DE, Niazy RK, Saunders J, Vickers J, Zhang Y, De Stefano N, Brady JM, Matthews PM (2004) Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage 23(Suppl 1):S208–S219 Good CD, Johnsrude IS, Ashburner J, Henson RN, Friston KJ, Frackowiak RS (2001) A voxel-based morphometric study of ageing in 465 normal adult human brains. Neuroimage 14(1 Pt 1):21–36. doi:10.1006/nimg 2001.0786 Jenkinson M, Bannister P, Brady M, Smith S (2002) Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage 17(2):825–841 Smith SM, Jenkinson M, Johansen-Berg H, Rueckert D, Nichols TE, Mackay CE, Watkins KE, Ciccarelli O, Cader MZ, Matthews PM, Behrens TEJ (2006) Tract-based spatial statistics: voxelwise
123
23.
24.
25.
26.
27.
28. 29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
analysis of multi-subject diffusion data. Neuroimage 31(4):1487– 1505 Eichler L, Bellenberg B, Hahn HK, Koster O, Schols L, Lukas C (2011) Quantitative assessment of brain stem and cerebellar atrophy in spinocerebellar ataxia types 3 and 6: impact on clinical status. AJNR Am J Neuroradiol 32(5):890–897. doi:10.3174/ajnr. A2387 Rub U, Brunt ER, Petrasch-Parwez E, Schols L, Theegarten D, Auburger G, Seidel K, Schultz C, Gierga K, Paulson H, van Broeckhoven C, Deller T, de Vos RA (2006) Degeneration of ingestion-related brainstem nuclei in spinocerebellar ataxia type 2, 3, 6 and 7. Neuropathol Appl Neurobiol 32(6):635–649. doi:10. 1111/j.1365-2990.2006.00772.x Rub U, Del Turco D, Del Tredici K, de Vos RA, Brunt ER, Reifenberger G, Seifried C, Schultz C, Auburger G, Braak H (2003) Thalamic involvement in a spinocerebellar ataxia type 2 (SCA2) and a spinocerebellar ataxia type 3 (SCA3) patient, and its clinical relevance. Brain 126(Pt 10):2257–2272. doi:10.1093/ brain/awg234 Asanuma C, Thach WT, Jones EG (1983) Distribution of cerebellar terminations and their relation to other afferent terminations in the ventral lateral thalamic region of the monkey. Brain Res 286(3):237–265 Jones DK, Horsfield MA, Simmons A (1999) Optimal strategies for measuring diffusion in anisotropic systems by magnetic resonance imaging. Magn Reson Med 42(3):515–525 Basser PJ, Mattiello J, Lebihan D (1994) MR diffusion tensor spectroscopy and imaging. Biophys J 66(1):259–267 Beaulieu C (2009) The biological basis of diffusion anisotropy. In: Johansen-Berg H, Behrens TEJ (eds) Diffusion MRI. Academic Press, Oxford, pp 105–126 Wheeler-Kingshott CA, Cercignani M (2009) About ‘‘axial’’ and ‘‘radial’’ diffusivities. Magn Reson Med 61(5):1255–1260. doi:10.1002/mrm.21965 Rub U, Gierga K, Brunt ER, de Vos RA, Bauer M, Schols L, Burk K, Auburger G, Bohl J, Schultz C, Vuksic M, Burbach GJ, Braak H, Deller T (2005) Spinocerebellar ataxias types 2 and 3: degeneration of the pre-cerebellar nuclei isolates the three phylogenetically defined regions of the cerebellum. J Neural Transm 112(11):1523–1545. doi:10.1007/s00702-005-0287-3 Yamada M, Tan CF, Inenaga C, Tsuji S, Takahashi H (2004) Sharing of polyglutamine localization by the neuronal nucleus and cytoplasm in CAG-repeat diseases. Neuropathol Appl Neurobiol 30(6):665–675. doi:10.1111/j.1365-2990.2004.00583.x Iwabuchi K, Tsuchiya K, Uchihara T, Yagishita S (1999) Autosomal dominant spinocerebellar degenerations. Clinical, pathological, and genetic correlations. Rev Neurol 155(4):255–270 Evert BO, Schelhaas J, Fleischer H, de Vos RA, Brunt ER, Stenzel W, Klockgether T, Wullner U (2006) Neuronal intranuclear inclusions, dysregulation of cytokine expression and cell death in spinocerebellar ataxia type 3. Clin Neuropathol 25(6):272–281 Durr A, Stevanin G, Cancel G, Duyckaerts C, Abbas N, Didierjean O, Chneiweiss H, Benomar A, Lyon-Caen O, Julien J, Serdaru M, Penet C, Agid Y, Brice A (1996) Spinocerebellar ataxia 3 and Machado-Joseph disease: clinical, molecular, and neuropathological features. Ann Neurol 39(4):490–499. doi:10.1002/ ana.410390411 Yamada M, Sato T, Tsuji S, Takahashi H (2008) CAG repeat disorder models and human neuropathology: similarities and differences. Acta Neuropathol 115(1):71–86. doi:10.1007/ s00401-007-0287-5 Yamada M, Hayashi S, Tsuji S, Takahashi H (2001) Involvement of the cerebral cortex and autonomic ganglia in Machado-Joseph disease. Acta Neuropathol 101(2):140–144 Matos CA, de Macedo-Ribeiro S, Carvalho AL (2011) Polyglutamine diseases: the special case of ataxin-3 and Machado-Joseph
J Neurol (2014) 261:291–299 disease. Prog Neurobiol 95(1):26–48. doi:10.1016/j.pneurobio. 2011.06.007 39. Johansen-Berg H, Behrens TEJ, Sillery E, Ciccarelli O, Thompson AJ, Smith SM, Matthews PM (2005) Functional-anatomical validation and individual variation of diffusion tractographybased segmentation of the human thalamus. Cereb Cortex 15(1):31–39 40. Klein JC, Barbe MT, Seifried C, Baudrexel S, Runge M, Maarouf M, Gasser T, Hattingen E, Liebig T, Deichmann R, Timmermann
299 L, Weise L, Hilker R (2012) The tremor network targeted by successful VIM deep brain stimulation in humans. Neurology 78(11):787–795. doi:10.1212/WNL.0b013e318249f702 41. Percheron G, Francois C, Talbi B, Meder JF, Fenelon G, Yelnik J (1993) The primate motor thalamus analysed with reference to subcortical afferent territories. Stereotact Funct Neurosurg 60(1– 3):32–41
123
