Grey matter volume changes over the whole brain in amyotrophic lateral sclerosis: A voxel-wise meta-analysis of voxel based morphometry studies

Transcription

1 Amyotrophic Lateral Sclerosis, 2010; 00: 1 6 ORIGINAL ARTICLE Grey matter volume changes over the whole brain in amyotrophic lateral sclerosis: A voxel-wise meta-analysis of voxel based morphometry studies ZHIYE CHEN & LIN MA Department of Radiology, Chinese PLA General Hospital, Beijing, China Abstract Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease with selected both upper and lower motor neuron involvement. Although some inconsistent results exist, both pathological studies and many structural neuroimaging studies have revealed brain volume changes in ALS. To provide an objective overview of structural changes in ALS, a voxel-wise meta-analysis was performed in published voxel based morphometry (VBM) studies. A systematic search of VBM studies was applied in ALS. Five studies met the inclusion criteria, comprising 84 ALS patients and 81 normal controls. A voxel-wise meta-analysis was performed on the retrieved VBM studies using signed differential mapping. Descriptive analysis showed that 25% of ALS patients had right precentral gyrus atrophy (2373 voxels). Group analysis demonstrated regional grey matter loss over the whole brain in the right precentral gyrus ( p ). Sensitivity analysis showed good sensitivity (157 voxels). In conclusion, right precentral grey matter atrophy was a common finding and prominent feature of brain structural changes in ALS. Key words: Amyotrophic lateral sclerosis, precentral gyrus, meta-analysis, voxel based morphometry Introduction Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease with selected both upper and lower motor neuron involvement (1,2). Pathological changes in cerebral cortical regions in ALS were confirmed as selective involvement of the motor cortex in a post mortem study (3). With the development of advanced computed neuroimaging, a novel automatic in vivo brain structural measurement technique, voxel-based morphometry (VBM) (4), has been developed, which provides a feasible and effective tool for brain structure research. In ALS, VBM studies showed that grey and white matter volume decreased widely, especially in the motor cortex (5,6), frontal and temporal regions (6 8), corpus callosum (9) and amygdala (10). However, one study found no volume reduction in primary motor cortices in ALS patients (11). These results were inconsistent to some extent and may be associated with many factors, including covariates, matching, sample size (12), image resolution (13), normalization template (14), threshold masking (15), imaging protocol (16), registration (17,18), full width half-maximum () (19,20) and statistical model (21). Other impact factors included some different clinical variables such as disease duration, cognitive state, and functional rating scores, etc. Although the inconsistency existed, individual VBM study revealed brain structural changes, especially subtle changes over the whole brain in ALS. Voxelwise meta-analysis could further integrate VBM findings or other functional neuroimaging studies, and obtain consistent results across different studies. Also, it has been used in obsessive-compulsive disorder (22), swallowing (23), emotional faces processing (24), etc. This analysis has been increasingly used to summarize functional neuroimaging results, and it can powerfully localize the active region or structural change region in neuroimaging studies. Meta-analysis methods mainly included kernel density analysis (KDA) (25), activation likelihood estimate (ALE) analysis (26), and multilevel kernel Correspondence: L. Ma, Department of Radiology, Chinese PLA General Hospital, 28 Fuxing Road, Beijing , China. cjr.malin@ vip.163.com (Received 14 June 2010 ; accepted 10 August 2010 ) ISSN print/issn X online 2010 Informa Healthcare DOI: /

2 2 Z. Chen & L. Ma density analysis (MKDA) (27). A novel voxel-wise meta-analysis technique, signed differential mapping (SDM) (22), is a quantitative meta-analysis method, combining various positive features of ALE and MKDA, and has been successfully used for metaanalysis in obsessive-compulsive disorder (28). To our knowledge, no voxel-wise meta-analysis on VBM studies in ALS has been reported. The aim of this study was to determine grey matter volume changes over the whole brain in ALS patients using SDM, and to identify possible specific aggregate changes in ALS. Materials and methods Search strategy A focused search strategy was designed as follows: in patients with ALS, brain structural changes over the whole brain were analyzed using voxel-based morphometry. The databases included Medline and PubMed. The first search expression included amyotrophic lateral sclerosis and motor neuron disease. The second search expression included voxel-based, voxel-wise, morphometry, and voxel based morphometry. This search strategy was performed on Medline and PubMed, and the operator AND was used between different concepts and operator OR was used between similar concepts. In addition, retrospective search and Google Scholar were also applied. Studies were included in this project if they met the following criteria: 1) VBM over the whole brain; 2) available peak coordinate for VBM; 3) sporadic ALS. Studies were excluded from the project if they were in line with exclusion criteria: 1) non-whole brain analysis; 2) not using VBM; 3) familial ALS; 4) not listing the coordinate in the article. Voxel-wise meta-analysis of VBM Voxel-wise meta-analysis was performed in the included VBM studies using signed differential mapping (SDM) ( (28). Descriptive analysis, group analysis and sensitivity analysis were carried out. In a pre-processing step, a relatively wide full width at half-maximum (, 25 mm) was used in order to account for different sources of spatial error, coregistration mismatch in the studies, and the size of the cluster or the location of the peak within the cluster. The statistical threshold was set to a p -value of without correction for false discovery rate (FDR). Results Included VBM studies Up to April 2010, a total of 52 articles were selected, including 39 articles without brain morphometry Table I. Demographic and clinical characteristics of VBM studies on ALS in meta-analysis. ALSFRS-R MMSE Education No. Age (years) Male (%) (years) Disease Subtype Onset duration ALS Controls ALS Controls ALS Controls ALS Controls (month) Definite Probable Possible Suspected Bulbar Limb Study Chang N/A 26.4 N/A N/A N/A Thivard N/A N/A N/A Ellis N/A N/A N/A N/A Mezzapesa N/A N/A N/A 26.2 Grosskreutz N/A N/A N/A ALS: amyotrophic lateral sclerosis; ALSFRS-R: ALS Functional Rating Scale-Revised; MMSE: Mini Mental State Examination; N/A: not available.

3 Grey matter volume changes over the whole brain in ALS 3 research and 13 articles with brain morphometry research. Of the 13 articles, five were included according to the inclusion criteria, and eight articles met the exclusion criteria (three not using VBM and five not listing the coordinate in the article). The included five VBM studies (6,11,29 31) included 84 ALS patients (mean age 51.6 years, F: M 7: 13) and 81 normal controls (mean age 51 years, F: M 1: 1). The ALS patients had mean disease duration of 27.5 months; 24 patients had bulbar onset and 60 patients had limb onset. The clinical details are shown in Table I. The technique details of the included VBM studies are described in Table II. MR scanners used in all the included studies were 1.5 T MR systems, and analysis software included Statistical Parametric Mapping 2 (SPM2) (four studies) and Analysis of Functional NeuroImages (AFNI) (one study). The VBM preprocessing procedure basically included three steps: normalization, segmentation, and smoothing. The smoothing kernel was 8 12 mm and one study ' s smoothing kernel was not available. The statistical threshold was p p , and one study p 0.05 (FDR corrected). Descriptive analysis All the included coordinates are shown in Figure 1. It showed that the significant clusters of the included VBM studies were mainly located in bilateral precentral gyri, frontal, temporal lobe, and left occipital and insula lobe. Descriptive analysis of quartiles was conducted with the included studies (Table III). Quartiles25 showed that 75% of studies detected atrophy of bilateral cingulate gyri and left inferior parietal lobule. Quartiles50 showed that 50% of studies detected right precentral gyrus atrophy (4795 voxels), right lentiform nucleus atrophy (68 voxels), and left middle frontal gyrus atrophy (6282 voxels). Quartiles75 showed that 25% of studies detected right precentral gyrus atrophy (two clusters, 2373 voxels), and other volume loss clusters located in the bilateral frontal lobe (especially right frontal lobe). From the descriptive analysis of quartiles, it was found that 25% 50% of studies detected right precentral gyrus atrophy in ALS patients. Group analysis (regional difference in grey matter over the whole brain) Group analysis was applied between ALS and control. Regional grey matter volume loss was mainly located in the right precentral gyrus (11 voxels, p , BA 4), and no other regions were found to have grey matter volume loss (Table IV, Figure 2). Sensitivity analysis Jackknife sensitivity analysis showed regional grey matter volume loss in right precentral gyrus (two clusters, 157 voxels) and left insula (one cluster, 18 voxels) (Table IV). The regional pattern of grey matter atrophy in the right precentral gyrus had a good reproducibility and it occurred in three of five included VBM studies. Discussion This meta-analysis used SDM (28) technique to summarize the results across published VBM studies on ALS. All the included VBM studies were papers published in peer-reviewed journals, followed the same inclusion criteria, and showed the peak cluster coordinates of grey matter volume deficit. The aim of this meta-analysis was to provide a grey matter volume deficit mapping of ALS compared with healthy controls. In this study, a novel voxel-wise meta-analysis technique was adopted to detect grey matter atrophy pattern. Compared with an individual VBM study, this technique established consistent VBM results from the included studies, and it also developed Table II. Technical details of VBM studies on ALS in meta-analysis. Study MRI scanner Software VBM process Smoothing p -value Chang Thivard Ellis Mezzapesa Grosskreutz SPM2 SPM2 AFNI Normalize to common template, segmented grey and white matter voxels value were multiplied by the Jacobian determinants Classic normalize, segment (with Hidden Markov Random Field approach), and smooth Using a computational algorithm to remove the extracerebral tissues, and manually segmenting the grey matter, white matter and CSF, then using AFNI software to infer difference in tissue distribution between groups 1.5 T (GE) SPM2 Generation of customized and prior probability maps (GM, WM, and CSF); then original images were normalized to the same stereotactic space and partitioned into GM, WM and CSF SPM2 Creating a customized template followed by normalization and segmentation with modulation of Jacobian determinants 12mm 10mm p (uncorrected) p 0.05 (FDR corrected) N/A p mm 8mm p (uncorrected) p (uncorrected)

4 4 Z. Chen & L. Ma Figure 1. The distribution graph of significant coordinate of all included VBM studies. a: axial, b: coronal, c: sagittal. some hypotheses derived from studies of braindamaged patients (32). SDM has been successfully used in the study of obsessive-compulsive disorder, and it has stricter selection criteria with regard to the reported peak coordinates and used a 25-mm adapted from that of ALE and preferred to that of MKDA (28). In this study, voxel-wise metaanalysis was conducted by using SDM, and the negative coordinates were used to reconstruct the signed differential mapping. Table III. Regional grey matter deficit of ALS at different quartiles. Quartiles Talairach coordinates SDM value Number of voxels Cluster location BA Quartiles25 Quartiles50 Quartiles75 60, 22, Left inferior parietal lobule 40 4, 28, Left posterior cingulate 23 6, 6, Right cingulate gyrus 24 36, 24, Right precentral gyrus 4 36,34, Left middle frontal gyrus 9 18, 4, Right lentiform nucleus Medial globus pallidus 36,24, Left middle frontal gyrus 9 24,40, Right superior frontal gyrus 9 50, 2, Right precentral gyrus 6 30, 6, Right middle frontal gyrus 6 40, 18, Right precentral gyrus 4 34,26, Right middle frontal gyrus 9 Continued voxels ( 10) was considered significant (SDM value 0.001). BA: Brodmann area.

5 Table IV. Group analysis and sensitivity analysis for VBM studies on ALS. Grey matter volume changes over the whole brain in ALS 5 Talairach coordinates SDM value p -value ( 10 4 ) Number of voxels Cluster location BA Group analysis Jackknife sensitivity analysis 36, 24, Right precentral gyrus 4 36, 24, Right precentral gyrus 4 36, 24, Right precentral gyrus 4 42,12, Left insula 13 Continued voxels ( 10) was considered significant (SDM value 0.001). BA: Brodmann area. By reviewing all the included VBM studies of ALS, it was found that ALS was not only a motor neuron disease (MND) but also a multisystem disease. ALS involved the bilateral precentral gyri as well as other brain regions. After descriptive analysis of quartiles, it was revealed that 25% 50% of VBM studies detected right precentral gyrus atrophy, and 25% detected frontal lobe (bilateral middle frontal and right superior frontal gyrus) atrophy. Group analysis further showed that only right precentral gyrus demonstrated atrophy, which should be a specific grey matter atrophy pattern for ALS according to this voxel-wise meta-analysis. Moreover, sensitivity analysis showed that right precentral gyrus atrophy had good sensitivity and reproducibility. Thus, right precentral gyrus atrophy may be a relatively common and specific finding in ALS, and more attention should be paid to this region in ALS patients, which may provide more information for clinical evaluation. Post mortem study (3) revealed the absence of Betz cells from layer 5 of the precentral cortex and the remaining pyramidal cells were significantly smaller than those seen in normal controls; in vivo brain structural volume changes could be effectively assessed by using VBM, and voxel-wise meta-analysis of VBM could be used to summarize the results of different laboratories and obtain consistent findings, and thus it should be more conclusive than individual VBM study. This meta-analysis study showed only the right precentral gyrus volume deficit, and this atrophy pattern presented asymmetry in ALS. The selective right precentral gyrus atrophy may be associated with neuronal degeneration in this area. Reasons for right precentral gyrus atrophy in ALS remain unknown, but the phenomenon is of great interest and further clinical study and structural asymmetry (33) study are needed to elucidate the underlying mechanism(s). A limitation of our study is that the number of included VBM studies was relatively small as ALS is an uncommon disease. Future studies needed to carry out further VBM studies should be of larger sample size. In summary, voxel-wise meta-analysis of VBM studies revealed that right precentral gyrus atrophy may be a specific grey matter volume deficit pattern for ALS. Figure 2. Regional grey matter volume loss in right precentral gyrus in ALS compared with normal control. The cluster representing atrophy of the right precentral gyrus (BA 4) is rendered on the template.

6 6 Z. Chen & L. Ma Acknowledgement This work was supported by the National Natural Science Foundation of China ( ). Declaration of interest : The authors have no conflicts of interest or disclosures. References 1. van der Graaff MM, de Jong JM, Baas F, de Visser M. Upper motor neuron and extra-motor neuron involvement in amyotrophic lateral sclerosis: a clinical and brain imaging review. Neuromuscul Disord 2009;19: Chou SM, Norris FH. Amyotrophic lateral sclerosis: lower motor neuron disease spreading to upper motor neurons. Muscle Nerve. 1993;16: Pringle CE, Hudson AJ, Munoz DG, Kiernan JA, Brown WF, Ebers GC. Primary lateral sclerosis: clinical features, neuropathology and diagnostic criteria. Brain. 1992;115: Ashburner J, Friston KJ. Voxel-based morphometry: the methods. Neuroimage. 2000;11: Kalra S, Arnold D. Neuroimaging in amyotrophic lateral sclerosis. Amyotroph Lateral Scler Other Motor Neuron Disord. 2003;4: Grosskreutz J, Kaufmann J, Fradrich J, Dengler R, Heinze HJ, Peschel T. Widespread sensorimotor and frontal cortical atrophy in amyotrophic lateral sclerosis. BMC Neurol. 2006;6: Abrahams S, Goldstein LH, Suckling J, Ng V, Simmons A, Chitnis X, Frontotemporal white matter changes in amyotrophic lateral sclerosis. J Neurol. 2005;252: Kassubek J, Unrath A, Huppertz HJ, Lule D, Ethofer T, Sperfeld AD, Global brain atrophy and corticospinal tract alterations in ALS, as investigated by voxel-based morphometry of 3-D MRI. Amyotroph Lateral Scler Other Motor Neuron Disord. 2005;6: Yamauchi H, Fukuyama H, Ouchi Y, Nagahama Y, Kimura J, Asato R, Corpus callosum atrophy in amyotrophic lateral sclerosis. J Neurol Sci. 1995;134: Pinkhardt E, van Elst L, Ludolph A, Kassubek J. Amygdala size in amyotrophic lateral sclerosis without dementia: an in vivo study using MRI volumetry. BMC Neurology. 2006; 6: Mezzapesa DM, Ceccarelli A, Dicuonzo F, Carella A, De Caro MF, Lopez M, Whole brain and regional brain atrophy in amyotrophic lateral sclerosis. AJNR Am J Neuroradiol. 2007;28: Pell GS, Briellmann RS, Chan CH, Pardoe H, Abbott DF, Jackson GD. Selection of the control group for VBM analysis: influence of covariates, matching and sample size. Neuroimage 2008;41: Pereira JM, Nestor PJ, Williams GB. Impact of inconsistent resolution on VBM studies. Neuroimage. 2008;40: Shen S, Szameitat AJ, Sterr A. VBM lesion detection depends on the normalization template: a study using simulated atrophy. Magn Reson Imaging. 2007;25: Ridgway GR, Omar R, Ourselin S, Hill DL, Warren JD, Fox NC. Issues with threshold masking in voxel-based morphometry of atrophied brains. Neuroimage. 2009;44: Tardif CL, Collins DL, Pike GB. Sensitivity of voxel-based morphometry analysis to choice of imaging protocol at 3 T. Neuroimage. 2009;44: Fein G, Landman B, Tran H, Barakos J, Moon K, Di Sclafani V, Statistical parametric mapping of brain morphology: sensitivity is dramatically increased by using brain-extracted images as inputs. Neuroimage. 2006;30: Bookstein FL. Voxel-based morphometry should not be used with imperfectly registered images. Neuroimage. 2001;14: Honea R, Crow TJ, Passingham D, Mackay CE. Regional deficits in brain volume in schizophrenia: a meta-analysis of voxel-based morphometry studies. Am J Psychiatry. 2005; 162: Wilke M, Kassubek J, Ziyeh S, Schulze-Bonhage A, Huppertz HJ. Automated detection of grey matter malformations using optimized voxel-based morphometry: a systematic approach. Neuroimage. 2003;20: Kawasaki Y, Suzuki M, Kherif F, Takahashi T, Zhou SY, Nakamura K, Multivariate voxel-based morphometry successfully differentiates schizophrenia patients from healthy controls. Neuroimage. 2007;34: Rotge JY, Guehl D, Dilharreguy B, Tignol J, Bioulac B, Allard M, Meta-analysis of brain volume changes in obsessive-compulsive disorder. Biol Psychiatry. 2009;65: Soros P, Inamoto Y, Martin RE. Functional brain imaging of swallowing: an activation likelihood estimation metaanalysis. Hum Brain Mapp. 2009;30: Fusar-Poli P, Placentino A, Carletti F, Landi P, Allen P, Surguladze S, Functional atlas of emotional faces processing: a voxel-based meta-analysis of 105 functional magnetic resonance imaging studies. J Psychiatry Neurosci. 2009;34: Wager TD, Smith EE. Neuroimaging studies of working memory: a meta-analysis. Cogn Affect Behav Neurosci. 2003;3: Laird AR, Fox PM, Price CJ, Glahn DC, Uecker AM, Lancaster JL, ALE meta-analysis: controlling the false discovery rate and performing statistical contrasts. Hum Brain Mapp. 2005;25: Wager TD, Lindquist MA, Nichols TE, Kober H, van Snellenberg JX. Evaluating the consistency and specificity of neuroimaging data using meta-analysis. Neuroimage. 2009;45: Radua J, Mataix-Cols D. Voxel-wise meta-analysis of grey matter changes in obsessive-compulsive disorder. Br J Psychiatry. 2009;195: Chang JL, Lomen-Hoerth C, Murphy J, Henry RG, Kramer JH, Miller BL, A voxel-based morphometry study of patterns of brain atrophy in ALS and ALS/FTLD. Neurology. 2005;65: Thivard L, Pradat PF, Lehericy S, Lacomblez L, Dormont D, Chiras J, Diffusion tensor imaging and voxel based morphometry study in amyotrophic lateral sclerosis: relationships with motor disability. J Neurol Neurosurg Psychiatry. 2007;78: Ellis CM, Suckling J, Amaro E Jr, Bullmore ET, Simmons A, Williams SC, Volumetric analysis reveals corticospinal tract degeneration and extramotor involvement in ALS. Neurology. 2001;57: Wager TD, Lindquist M, Kaplan L. Meta-analysis of functional neuroimaging data: current and future directions. Soc Cogn Affect Neurosci. 2007;2: Watkins KE, Paus T, Lerch JP, Zijdenbos A, Collins DL, Neelin P, Structural asymmetries in the human brain: a voxel-based statistical analysis of 142 MRI scans. Cereb Cortex. 2001;11: