Atormac
briv
Neurology India
menu-bar5 Open access journal indexed with Index Medicus
  Users online: 449  
 Home | Login 
About Editorial board Articlesmenu-bullet NSI Publicationsmenu-bullet Search Instructions Online Submission Subscribe Videos Etcetera Contact
  Navigate Here 
 Search
 
  
 Resource Links
  »  Similar in PUBMED
 »  Search Pubmed for
 »  Search in Google Scholar for
 »Related articles
  »  Article in PDF (2,647 KB)
  »  Citation Manager
  »  Access Statistics
  »  Reader Comments
  »  Email Alert *
  »  Add to My List *
* Registration required (free)  

 
  In this Article
 »  Abstract
 » Introduction
 »  Materials and Me...
 » Results
 » Discussion
 » Conclusion
 »  References
 »  Article Figures
 »  Article Tables

 Article Access Statistics
    Viewed318    
    Printed6    
    Emailed0    
    PDF Downloaded13    
    Comments [Add]    

Recommend this journal

 


 
Table of Contents    
ORIGINAL ARTICLE
Year : 2021  |  Volume : 69  |  Issue : 5  |  Page : 1318-1325

Cortical and Subcortical Brain Area Atrophy in SCA1 and SCA2 Patients in India: The Structural MRI Underpinnings and Correlative Insight Among the Atrophy and Disease Attributes


1 Department of Zoology, Nalbari College, Assam, India
2 Department of Physiology, VMMC & Safdarjung Hospital, New Delhi, India
3 Department of Computational Biology, Indraprastha Institute of Information Technology, Delhi, India
4 Department of Biomedical Engineering, All India Institute of Medical Sciences; Center for Biomedical Engineering, Indian Institute of Technology Delhi, New Delhi, India
5 Genomics and Molecular Medicine, CSIR-Institute of Genomics and Integrative Biology (CSIR -IGIB), New Delhi, India
6 Department of Physiology, All India Institute of Medical Sciences, New Delhi, India
7 Department of Neurology, All India Institute of Medical Sciences, New Delhi, India
8 Department of NMR, All India Institute of Medical Sciences, New Delhi, India

Date of Submission18-Aug-2019
Date of Decision04-Dec-2019
Date of Acceptance05-Feb-2020
Date of Web Publication30-Oct-2021

Correspondence Address:
Kishore K Deepak
Head, Department of Physiology Room No. 2009, Teaching Block, Second Floor, New Delhi-110 029
India
Login to access the Email id

Source of Support: None, Conflict of Interest: None


DOI: 10.4103/0028-3886.329596

Rights and Permissions

 » Abstract 


Introduction: Genetically defined spinocerebellar ataxia (SCA) type 1 and 2 patients have differential clinical profile along with probable distinctive cortical and subcortical neurodegeneration. We compared the degree of brain atrophy in the two subtypes with their phenotypic and genotypic parameters.
Methods: MRI was performed using a 3T scanner (Philips, Achieva) to obtain 3D T1-weighted scans of the whole brain and analyzed by FreeSurfer (version 5.3 and 6 dev.) software. Genetically proven SCA1 (n = 18) and SCA2 (n = 25) patients with age-matched healthy controls (n = 8) were recruited. Clinical severity was assessed by the International Cooperative Ataxia Rating Scale (ICARS). To know the differential pattern of atrophy, the groups were compared using ANOVA/Kruskal-Wallis test and followed by correlation analysis with multiple corrections. Further, machine learning-based classification of SCA subtypes was carried out.
Result: We found (i) bilateral frontal, parietal, temporal, and occipital atrophy in SCA1 and SCA2 patients; (ii) reduced volume of cerebellum, regions of brain stem, basal ganglia along with the certain subcortical areas such as hippocampus, amygdala, thalamus, diencephalon, and corpus callosum in SCA1 and SCA2 subtypes; (iii) higher subcortical atrophy SCA2 than SCA1 (iv) correlation between brain atrophy and disease attributes; (v) differential predictive pattern of two SCA subtypes using machine learning approach.
Conclusion: The present study suggests that SCA1 and SCA2 do not differ in cortical thinning while a characteristic pattern of subcortical atrophy SCA2 > SCA1 is observed along with correlation of brain atrophy and disease attributes. This may provide the diagnostic guidance of MRI to SCA subtypes and differential therapies.


Keywords: Correlation, cortical thinning, FreeSurfer, machine learning, spinocerebellar ataxia, subcortical atrophy
Key Message: The brain area atrophy of cortical and subcortical regions has been seen in SCA1 and SCA2 patients. The subcortical atrophy was found to be more in SCA2 than SCA1 patients which would have a diagnostic value.


How to cite this article:
Tamuli D, Kaur M, Sethi T, Singh A, Faruq M, Jaryal AK, Srivastava AK, Kumaran SS, Deepak KK. Cortical and Subcortical Brain Area Atrophy in SCA1 and SCA2 Patients in India: The Structural MRI Underpinnings and Correlative Insight Among the Atrophy and Disease Attributes. Neurol India 2021;69:1318-25

How to cite this URL:
Tamuli D, Kaur M, Sethi T, Singh A, Faruq M, Jaryal AK, Srivastava AK, Kumaran SS, Deepak KK. Cortical and Subcortical Brain Area Atrophy in SCA1 and SCA2 Patients in India: The Structural MRI Underpinnings and Correlative Insight Among the Atrophy and Disease Attributes. Neurol India [serial online] 2021 [cited 2021 Dec 3];69:1318-25. Available from: https://www.neurologyindia.com/text.asp?2021/69/5/1318/329596





 » Introduction Top


Spinocerebellar ataxia (SCA) is a group of neurodegenerative disorders among which SCA1 and SCA2 are prevalent subtypes worldwide.[1],[2] SCA subtypes are highly heterogeneous based on their clinical signs and symptoms including gait imbalance, ophthalmoplegia, dysarthria, and pyramidal and extrapyramidal signs. The clinical diversity is a consequence of unstable CAG trinucleotide repeat expansion at ATXN1 and ATXN2 genes for SCA1 and SCA2 mutants, respectively. The expanded CAG triplets lead to abnormal polyQ stretches in the encoded proteins, which, in turn, recruit other expanded molecules of ataxin-1/ataxin-2 to assemble in protein aggregates leading to neurodegeneration.[1] It could be indicative of a differential neuronal loss due to genetic mutation in SCA subtypes and may be responsible for the differential clinical profile. Thus, genotype-phenotype correlates are important to delineate the underlying pathophysiology. Previous studies have found an association of CAG repeat length with clinical scoring and age at onset.[1],[2],[3],[4],[5],[6]

Neurodegeneration forms the primary pathophysiological feature of SCA patients. Therefore, to characterize the degree of neuronal loss of brain in SCA patients, neuroimaging studies have been widely utilized. Magnetic resonance imaging (MRI) is a commonly used neuroimaging technique to assess the status of brain damage in SCA patients. However, the characterization of SCA1 and SCA2 based on differential neuronal loss has been poorly documented. Subcortical brain area atrophy particularly cerebellum and its direct connections like brainstem and basal ganglia were mainly investigated as of their prominent role in ataxia. Most of the reported literature have used semiautomated morphometric and volumetric MRI analysis to quantify subcortical atrophy in SCA subtypes.[7],[8],[9],[10] While subcortical atrophy forms the basis of most clinical symptoms, patients also suffer from cognitive dysfunction.[11] However, the comprehensive evaluation of cortical involvement in SCA1 and SCA2 is lacking. Only a few studies have reported cortical atrophy along with subcortical neurodegeneration in SCA1 and SCA2 using automated voxel-based morphometry (VBM) analysis of MRI.[12],[13] However, due to the broad overlapping among the SCA subtypes, a very sensitive analytical approach is rather important. Remarkably, the automated unbiased FreeSurfer software utilizes a surface-based model which is more sensitive to identify cerebral cortex abnormalities as compared to VBM.[14] Therefore, we carried out an extensive investigation of the cortical and subcortical atrophy in SCA1 and SCA2 patients using the FreeSurfer software for structural MRI analysis. A secondary objective was to evaluate the correlations of brain areas with the clinical and genetic profile of SCA patients to elucidate the pathophysiological basis of the disease severity. We also tried to predict the SCA subtypes by a machine learning approach based on brain area atrophy. This may help in diagnostic guidance of MRI to SCA subtypes by clinicians for rational therapies.


 » Materials and Methods Top


We enrolled 43 patients with SCA (SCA1 = 18 and SCA2 = 25) from the outpatient Department of Neurology at All India Institute of Medical Sciences (AIIMS), New Delhi between March 2014 and January 2016. All recruited patients fulfilled the inclusion criteria of positive genetic confirmation for either SCA1 or SCA2 without having MRI contraindications (any neural, cardiac, or cochlear implant, or history of claustrophobia). Eight age-matched healthy individuals were recruited as controls. Informed written consent was obtained according to the Declaration of Helsinki from all research participants. The study protocol was approved by the Institutional Ethics Committee (IESC/T-447/29.11.2013), AIIMS, New Delhi.

After the genetic testing, all recruited patients underwent a clinical evaluation (performed by neurologist) followed by neuroimaging study. The same protocol of MRI procedure was performed in 8 healthy controls.

Genetic analysis

The size of CAG repeat length was assessed by fragment analysis tool in GeneMapper software v. 4 (ABI) and confirmed by Sanger sequencing method using PCR.

Clinical evaluation

Clinical evaluation of each SCA patient was performed by using the International Cooperative Ataxia Rating Scale (ICARS) - a validated score to quantify cerebellar ataxia by the intensity of the symptoms.[15]

MRI data acquisition

All participants went through an extensive MRI protocol using 3T scanner (Philips, Achieva) and 32 channel head coil to obtain 3D T1-weighted scans of the whole brain with following protocol: slice thickness of 1 mm, echo time (TE) =3.2 ms, repetition time (TR) =7.1 ms, flip angle = 8°, voxel size = 0.6 × 1 × 1 mm3 (acquired); 0.2 × 0.2 × 1 mm3 (reconstructed) and field-of-view (FOV) =240 × 240 × 180 mm3. T2-weighted scans were also acquired - TE = 30 ms, TR = 3300 ms, voxel size = 0.5 × 0.5 × 1.1 mm3 (reconstructed) and FOV = 250 × 250 × 165 mm3. The images with significant motion artifact were excluded. Later, the T1-weighted images were used for both cortical and subcortical analysis by FreeSurfer software (ver. 5.3 and 6 dev.).

FreeSurfer analysis pipeline

The whole-brain segmentation was performed using FreeSurfer version 5.3 (http://surfer.nmr.mgh.harvard.edu/). This includes the surface-based analysis of cerebral cortex and volumetric analysis of subcortical brain areas. The FreeSurfer analysis pipeline was carried out on T1-weighted images for whole-brain segmentation. The stepwise processes of automated sequences under the pipeline were performed according to the protocol described by Fischl.[16]

In short, motion correction and removal of non-brain tissues from the image were followed by automated Talairach transformation.[17] Further, subcortical white matter and deep gray matter volumetric structures were segmented,[18] followed by stepwise procedures of intensity normalization,[19] tessellation of the gray matter-white matter boundary and the gray matter-CSF boundary, and automated topology correction.[20],[21] Finally, cortical modeling was done by surface deformation following intensity gradients to optimally place gray/white and gray/cerebrospinal fluid borders at the location where the greatest shift in intensity defines the transition to the other tissue class.[22],[23] Cortical thickness was calculated as the closest distance from the gray/white boundary to the gray/cerebrospinal fluid boundary at each vertex on the tessellated surface.[23]

Cortical thickness and subcortical volume calculation

Cortical thickness (mm) was calculated with FreeSurfer's automatic parcellation of the cerebral cortex into 68 regions (34 per hemisphere). FreeSurfer software automatically segments the subcortical region to quantify subcortical volume. After that, the brainstem was segmented into the midbrain, pons, medulla oblongata, and superior cerebellar peduncle by using the brainstem module in the development version of FreeSurfer (ver. 6; ftp://surfer.nmr.mgh.harvard.edu/pub/dist/freesurfer/dev/).[24]

Statistical maps were produced by using FreeSurfer's Query, Design, Estimate, Contrast (QDEC) interface (https://surfer.nmr.mgh.harvard.edu/fswiki/FsTutorial/QdecGroupAnalysis_ Freeview), thus yielding group averages and inference on the cortical morphometric data produced by the FreeSurfer processing stream [Figure 1].
Figure 1: Map of group difference between SCA subtypes and controls (SCA1 vs. controls and SCA2 vs. controls) for cortical thickness; a general linear model (GLM) was computed vertex-by-vertex for this analysis. The colour bar represents the cortical thickness - more towards the negative side or light blue indicates higher cortical thinning

Click here to view


Statistical analysis

Each parameter was tested for data distribution based on standard normality tests (D' Agostino-Pearson omnibus normality test and Shapiro–Wilk test). Data with normally distributed parameters were expressed as the mean ± SD and nonparametric data were expressed as median with the interquartile range using Graph-Pad Prism Version 5.00 for Windows (GraphPad Software, Inc., USA). One-way ANOVA with Tukey's post hoc multiple comparison test for normally distributed data and Kruskal-Wallis test with Dunn's post hoc multiple comparisons for non-Gaussian distribution were used. The two-tailed level of statistical significance was set at P < 0.05. The categorical variables were compared by using Fisher's exact test [Table 1]. Also, cortical thickness and subcortical volume of patient/control discrimination for each anatomical structure were separately assessed with Cohen's d in IBM-SPSS ver. 24 [Supplementary Table S1],[Supplementary Table S2],[Supplementary Table S3]. Association between continuous variables was determined by Pearson and Spearman correlation analysis based on the data distribution by Graph-Pad Prism Version 5.00. Multiple corrections for correlation was done using the Benjamini-Hochberg method where the false discovery rate was 0.25.
Table 1: Demographic and disease characteristics (clinical and genetic) of the study population

Click here to view



Dimensionality reduction and feature selection were done using a battery of standard statistical and machine learning models - Chi-square, information gain, entropy-based methods, and Boruta algorithm.[25] Data were divided into training and validation datasets. The training dataset was 2/3 of the dataset while the validation dataset was the remaining 1/3 of the dataset. The Random Forest[26] model was trained for classification of SCA1 and SCA2. The model accuracies were validated on the validation set.


 » Results Top


SCA patients and healthy controls

[Table 1] summarizes the demographic and disease characteristics of the study population. SCA subtypes and controls were comparable in age and male: female distribution pattern. No statistical difference was observed in age at onset and disease duration amongst patient groups. SCA subgroups were classified based on CAG trinucleotide repeats and it lies on the same range as previously reported in the literature.[6],[10] The clinical dysfunction assessed by ICARS was similar in SCA1 and SCA2.

Surface-based cortical thickness analysis

Cortical thinning was observed in the patients of each SCA subtype as compared with healthy controls but no statistically significant difference was found in the comparisons made among the two SCA subtypes [Table 2] and [Table 3], [Figure 1]. Indeed, significant cortical thinning was observed in all the four lobes of the cerebral cortex (frontal, parietal, temporal and occipital) of both left and right hemispheres in SCA1 and SCA2 patients as compared to controls. Similarly, the thinning of certain cortical areas like cingulate of both the hemispheres was measured while the thinning of the insula was observed only in the right hemisphere in SCA1 and SCA2 than controls. Thus, the composite mean cortical thickness of both the hemispheres was significantly reduced in SCA1 and SCA2 patients than controls [Table 2] and [Table 3].
Table 2: Comparison of the cortical thickness of left cerebral hemisphere in the two SCA subtypes along with healthy controls

Click here to view
Table 3: Comparison of the cortical thickness of right cerebral hemisphere in the two SCA subtypes along with healthy controls

Click here to view


Volumetric subcortical analysis

Interestingly, we found volumetric subcortical atrophy in SCA subtypes than controls along with the differential reduction of volume in the cerebellum, brain stem, and corpus callosum among the SCA1 and SCA2 [Table 4]. The volume of different structures of the cerebellum, brain stem, basal ganglia (caudate, putamen, pallidum and accumbens), and certain subcortical structures such as the thalamus, hippocampus, amygdala, and right ventral diencephalon were significantly decreased in the SCA1 and SCA2 patients than controls. We have also seen the characteristic feature of neurodegenerative diseases - significant enlargement (P = <0.001) of the volume of 4th ventricle in SCA1 (2608.85 (2501.80– 3352.45) mm3) and SCA2 (3522.30 (2721.45–4765.95) mm3) than controls (1504.30 (1189.60–1667.70 mm3).
Table 4: Volume of subcortical brain structures in the two subtypes of spinocerebellar ataxia patients compared with healthy controls

Click here to view


In the case of differential reduction, the volume of the bilateral cerebellar cortex and cerebellar white matter were significantly decreased in SCA2 than SCA1. The volumetric atrophy was calculated in the structures of the brain stem - midbrain, pons, medulla as well as the total brainstem along with the superior cerebellar peduncle in all the three SCA subtypes versus healthy controls [Supplementary Figure S1]. The midbrain, pontine, and total brainstem atrophy was more in SCA2 than SCA1 patients. Furthermore, the loss of volume in the right ventral diencephalon, corpus callosum central, and corpus callosum mid posterior were found to be higher in SCA2 as compared to SCA1 patients. Therefore, the overall cerebellar atrophy in a decreasing trend was SCA2 > SCA1. Thus, in the subcortical volumetric analysis - SCA2 subtype emerges to be more severely affected.



Furthermore, to know the difference of neurodegeneration in patient and control groups rather than confounding this with sample size, we quantified the brain area atrophy based on effect size calculation in SCA1, SCA2, and control groups [Supplementary Table S1],[Supplementary Table S2],[Supplementary Table S3].

Correlation analysis

Correlation between brain area atrophy and disease characteristics

We conducted a correlation analysis between the brain areas and disease characteristics such as clinical severity measured by ICARS, CAG repeat length, disease duration, and age at onset.

The substantial inverse correlation was found among the ICARS and brain areas (cortical thickness and subcortical volume) in both SCA subtypes. In SCA1, midbrain (r = -0.713, P = 0.001), pons (r = -0.850, P < 0.001), medulla (r = -0.720, P = 0.001), and brainstem (r = -0.832, P < 0.001) were inversely correlated with ICARS [Supplementary Table S4]. Interestingly, all four lobes of the cerebral cortex (frontal, parietal, temporal, and occipital), cerebellar cortex and pons were negatively associated with clinical severity in SCA2 patients [Supplementary Table S5].



None of the brain areas was found to be correlated with CAG repeat length in SCA1 and SCA2 patients.

The midbrain (r = -0.711, P = 0.001), pons (r = -0.802, P < 0.001), and brainstem (r = -0.761, P < 0.001) were inversely associated with disease duration in SCA1 patients [Supplementary Table S6]. While no correlation was calculated among the disease duration and brain areas in SCA2 patients.



Similarly, the age at onset was not correlated with the brain areas in SCA1 and SCA2.

Correlation among the parameters of disease characteristics

We found a significant positive association between ICARS and disease duration (r = 0.567, P < 0.001) while significant negative association among the CAG repeat length and age at onset (r = -0.705, P = 0.001) in SCA1 patients [Supplementary Figure S2]. In SCA2 patients, ICARS was significantly associated with disease duration (r = 0.449, P = 0.024) and remarkably the genotype-phenotype correlation was seen where a significant correlation of ICARS and CAG repeat length (r = 0.408, P = 0.043) was found [Supplementary Figure S3]. The CAG repeat length was inversely associated with age at onset (r = -0.759, P < 0.001) in SCA2. [Supplementary Figure S3].



Machine learning-based classification of SCA subtypes

Since we found brain areas to be differentially involved in SCA1 and SCA2 patients, we confirmed if these differences could be leveraged to construct predictive models for SCA subtypes. A random forest classifier built with the cortical and subcortical brain areas (left pericalcarine, right ventral diencephalon, corpus callosum central, left vessel, right lateral ventricle, left cerebellar white matter, right cerebellar white matter, and brainstem) achieved an overall accuracy of 85.7% in classifying SCA1 from SCA2 [Supplementary Table S7]. SCA1 had 100% sensitivity and 80% specificity while SCA2 exhibited 80% sensitivity and 100% specificity.




 » Discussion Top


To our knowledge, this is the first study that reports cortical and subcortical brain area atrophy in SCA1 and SCA2 by surface-based analysis approach using FreeSurfer software in the Indian scenario. Comparable cortical thinning (frontal, parietal, temporal, and occipital lobes) in both SCA while differential subcortical atrophy (bilateral cerebellar cortex as well as cerebellar white matter along with the midbrain, pons, brainstem, ventral diencephalon, and corpus callosum) in SCA2 than SCA1 (SCA2 > SCA1).

Reported literature primarily mentioned subcortical atrophy in SCA1 and SCA2 by semiautomated MRI study.[7],[9],[10] They found cerebellar and brainstem atrophy in both SCA while one study found higher atrophy in SCA2 than SCA1.[10] Very few studies have reported neurodegeneration of cerebral cortex in SCA1 and SCA2. Della Nave et al. found the cortical and subcortical atrophy in SCA1 and SCA2 by using automated voxel-based morphometry (VBM) analysis which is diagnostically better (unbiased) than previously mentioned semiautomated studies.[13] They quantified gray and white matter atrophy in SCA1 involving all four lobes of the cerebral cortex and cerebellar and brain stem atrophy in accordance with our results. Nevertheless, in contrast, only frontal, parietal, and cingular atrophy was seen in SCA2. Distinctively, SCA1 and SCA2 were found to be comparable. The difference with our study lies in lower sample size (SCA1 = 10, SCA2 = 10) and analysis approach.

Moreover, Goel et al., 2011 have revealed cortical and subcortical brain atrophy in SCA subtypes using optimized VBM analysis of MRI.[12] SCA1 patients were found to have left inferior frontal gyrus and right superior temporal gyrus atrophy with the significantly atrophic white matter in the bilateral frontal lobe. SCA2 patients had bilateral superior temporal gyri atrophy with white matter atrophy in bilateral frontal and temporal lobes and precuneus when compared to the control group. Evidently, the pontocerebellar atrophy has been seen in both SCA subtypes with greater pontine atrophy in SCA2 than SCA1 similar to our study. Besides, they found no significant difference in the gray matter atrophy among SCA1 and SCA2. However, the involvement of frontal and temporal in SCA1 and frontal, temporal, and parietal in SCA2 sparing the other lobes of cerebral cortex differ from our study. This disparity may ascribe to the usage of software for MRI analysis. In our study, we utilized the current gold standard surface-based analysis approach for cortical thickness analysis using FreeSurfer software.[14] It uses exact geometry of cortical surface to do intersubject registration and thus, providing a better matching of homologous cortical regions with greater accuracy.[27] Using FreeSurfer, we have been able to assess comprehensively the degree of cortical atrophy in SCA1 and SCA2.

On correlation analysis, we found a significant negative correlation among ICARS and brain structures. The midbrain, pons, medulla, and brainstem inversely correlated with ICARS in SCA1 while brain areas of all four lobes of cerebral cortex along with the pons and cerebellar cortex were negatively associated with SCA2 patients. Thus, ICARS predicts the disease severity analyzed by FreeSurfer – severe the disease, higher brain atrophy. Similar findings of negative associations were reported between ICARS and - cerebellum and brainstem in both SCA[13]; cerebellum in SCA1[12]; brainstem and pons in SCA1[28]; and cerebellum in SCA2.[29]

No brain areas were correlated with the CAG repeat length in both SCA suggesting no influence of genotype on brain area atrophy as mentioned by Schols et al.[1] in SCA. In addition, we found a significant negative association of disease duration with the regions of the brainstem in SCA1 patients. This could imply that a longer duration of SCA1 leads to more neuronal loss because of defects in axonal transport.[1]

In line with the earlier literature,[3],[4],[5] we have reported a significant negative correlation between age at onset and CAG repeat suggesting genetic anticipation in SCA1 and 2 patient groups. Similar to the previous report,[6] we have found a significant positive correlation between ICARS and CAG repeat length in SCA2 subtype defines the genotype-phenotype correlation. Similarly, a significant positive correlation was found between ICARS and disease duration in both SCA1 and SCA2 thus, the longer duration would result in more severe clinical symptoms.[6]

Interestingly, the present study is the first one to suggest MR volumetry-based classification models that could differentiate and predict SCA1 or SCA2 using a handful features of brain atrophy selected by machine learning approach as mentioned in our earlier report.[30] Such analyses may provide quick decision support to neurologists in difficult to recognize cases without the additional cost of genetic testing. Although the inclusion of a higher number of SCA patients in each subtype (especially SCA3) would have aided in a better demarcation between the two subtypes.


 » Conclusion Top


The findings of this study suggest that SCA1 and SCA2 have a higher degree of cortical and subcortical atrophy than controls. Both SCA does not differ in cortical thinning but in subcortical atrophy, an intriguing pattern of SCA2 > SCA1 emerges. Clinical severity (ICARS) correlates with neurodegeneration and genetic anticipation was found in both SCA subtypes. The genotype-phenotype correlation was observed in SCA2 but no association among the genotype and neurodegeneration. We have also suggested a novel machine learning approach to segregate SCA1 and SCA2 based on brain area involvement.

Acknowledgements

We acknowledge the Indian Council of Medical Research (ICMR), New Delhi, India for providing Research Associateship (45/9/2018-PHY/BMS) for the author D. Tamuli.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.



 
 » References Top

1.
Schöls L, Bauer P, Schmidt T, Schulte T, Riess O. Autosomal dominant cerebellar ataxias: Clinical features, genetics, and pathogenesis. Lancet Neurol 2004;3:291-304.  Back to cited text no. 1
    
2.
Tamuli D, Kaur M, Boligarla A, Jaryal AK, Srivastava AK, Deepak KK. Depressed baroreflex sensitivity from spontaneous oscillations of heart-rate and blood-pressure in SCA1 and SCA2. Acta Neurol Scand 2019. doi: 10.1111/ane. 13151.  Back to cited text no. 2
    
3.
Cancel G, Dürr A, Didierjean O, Imbert G, Bürk K, Lezin A, et al. Molecular and clinical correlations in spinocerebellar ataxia 2: A study of 32 families. Hum Mol Genet 1997;6:709-15.  Back to cited text no. 3
    
4.
Yeh T-H, Lu C-S, Chou Y-HW, Chong C-C, Tony Wu N-HH, Chen R-S. Autonomic dysfunction in Machado-Joseph disease. Arch Neurol 2005;62:630-6.  Back to cited text no. 4
    
5.
Pradhan C, Yashavantha BS, Pal PK, Sathyaprabha TN. Spinocerebellar ataxias type 1, 2 and 3: A study of heart rate variability. Acta Neurol Scand 2008;117:337-42.  Back to cited text no. 5
    
6.
Montes-Brown J, Sánchez-Cruz G, García AM, Báez ME, Velázquez-Pérez L. Heart rate variability in type 2 spinocerebellar ataxia. Acta Neurol Scand 2010;122:329-35.  Back to cited text no. 6
    
7.
Guerrini L, Lolli F, Ginestroni A, Belli G, Della Nave R, Tessa C, et al. Brainstem neurodegeneration correlates with clinical dysfunction in SCA1 but not in SCA2. A quantitative volumetric, diffusion and proton spectroscopy MR study. Brain 2004;127:1785-95.  Back to cited text no. 7
    
8.
Nagaoka U, Suzuki Y, Kawanami T, Kurita K, Shikama Y, Honda K, et al. Regional differences in genetic subgroup frequency in hereditary cerebellar ataxia, and a morphometrical study of brain MR images in SCA1, MJD and SCA6. J Neurol Sci 1999;164:187-94.  Back to cited text no. 8
    
9.
Bürk K, Abele M, Fetter M, Dichgans J, Skalej M, Laccone F, et al. Autosomal dominant cerebellar ataxia type I clinical features and MRI in families with SCA1, SCA2 and SCA3. Brain 1996;119:1497-505.  Back to cited text no. 9
    
10.
Klockgether T, Skalej M, Wedekind D, Luft AR, Welte D, Schulz JB, et al. 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 1998;121:1687-93.  Back to cited text no. 10
    
11.
Bürk K, Globas C, Bösch S, Klockgether T, Zühlke C, Daum I, et al. Cognitive deficits in spinocerebellar ataxia type 1, 2, and 3. J Neurol 2003;250:207-11.  Back to cited text no. 11
    
12.
Goel G, Pal PK, Ravishankar S, Venkatasubramanian G, Jayakumar PN, Krishna N, et al. Gray matter volume deficits in spinocerebellar ataxia: An optimized voxel based morphometric study. Park Relat Disord 2011;17:521-7.  Back to cited text no. 12
    
13.
Della Nave R, Ginestroni A, Tessa C, Salvatore E, De Grandis D, Plasmati R, et al. Brain white matter damage in SCA1 and SCA2. An in vivo study using voxel-based morphometry, histogram analysis of mean diffusivity and tract-based spatial statistics. Neuroimage 2008;43:10-9.  Back to cited text no. 13
    
14.
De Rezende TJ, D'abreu A, Guimaraes RP, Lopes TM, Lopes-Cendes I, Cendes F, et al. Cerebral cortex involvement in Machado-Joseph disease. Eur J Neurol 2015;22:277-83.  Back to cited text no. 14
    
15.
Trouillas P, Takayanagi T, Hallett M, Currier RD, Subramony SH, Wessel K, et al. InternationalCooperativeAtaxia Rating Scale for pharmacological assessmentof the cerebella syndrome. J Neurol Sci 1997;145:205-11.  Back to cited text no. 15
    
16.
Fischl B. FreeSurfer. Neuroimage 2012;62:774-81.  Back to cited text no. 16
    
17.
Talairach J, Tournoux P., “Co planar stereotaxic atlas of the human brain: 3 dimensional proportional system”, An Approach to Cerebral Imaging. 1988.  Back to cited text no. 17
    
18.
Fischl B, Salat DH, Busa E, Albert M, Dieterich M, Haselgrove C, et al. Whole brain segmentation: Neurotechnique automated labeling of neuroanatomical structures in the human brain. Neuron 2002;33:341-55.  Back to cited text no. 18
    
19.
Sled JG, Zijdenbos AP, Evans AC. A nonparametric method for automatic correction of intensity nonuniformity in MRI data. IEEE Trans Med Imaging 1998;17:87-97.  Back to cited text no. 19
    
20.
Fischl B, Liu A, Dale AM. Automated manifold surgery: Constructing geometrically accurate and topologically correct models of the human cerebral cortex. IEEE Trans Med Imaging 2001;20:70-80.  Back to cited text no. 20
    
21.
Ségonne F, Pacheco J, Fischl B. Geometrically accurate topology-correction of cortical surfaces using nonseparating loops. IEEE Trans Med Imaging 2007;26:518-29.  Back to cited text no. 21
    
22.
Dale AM, Fischl B, Sereno MI. Cortical surface-based analysis. Neuroimage 1999;9:179-94.  Back to cited text no. 22
    
23.
Fischl B, Dale AM. Measuring the thickness of the human cerebral cortex from magnetic resonance images. Proc Natl Acad Sci 2000;97:11050-5.  Back to cited text no. 23
    
24.
Iglesias JE, Van Leemput K, Bhatt P, Casillas C, Dutt S, Schuff N, et al. Bayesian segmentation of brainstem structures in MRI. Neuroimage 2015;113:184-95.  Back to cited text no. 24
    
25.
Kursa MB, Rudnicki WR. Feature selection with the Boruta package. J Stat Softw 2010;36:1-13.  Back to cited text no. 25
    
26.
Breiman L. Random forests. Mach Learn 2001;45:5-32.  Back to cited text no. 26
    
27.
Greve DN. An absolute beginner's guide to surface- and voxel-based morphometric analysis. Proc Int Soc Magn Reson Med 2011;i: 1-7.  Back to cited text no. 27
    
28.
Schulz JB, Borkert J, Wolf S, Schmitz-Hübsch T, Rakowicz M, Mariotti C, et al. Visualization, quantification and correlation of brain atrophy with clinical symptoms in spinocerebellar ataxia types 1, 3 and 6. Neuroimage 2010;49:158-68.  Back to cited text no. 28
    
29.
Brenneis C, Bösch SM, Schocke M, Wenning GK, Poewe W. Atrophy pattern in SCA2 determined by voxel-based morphometry. Neuroreport 2003;14:1799-802.  Back to cited text no. 29
    
30.
Tamuli D, Godiyal AK, Kaur M, Jaryal AK, Srivastava AK, Deepak KK. Autonomic function based classification of spinocerebellar ataxia type 1 and 2 using machine learning classifiers. Indian J Physiol Pharmacol 2017;61;323-32.  Back to cited text no. 30
    


    Figures

  [Figure 1]
 
 
    Tables

  [Table 1], [Table 2], [Table 3], [Table 4]



 

Top
Print this article  Email this article
   
Online since 20th March '04
Published by Wolters Kluwer - Medknow