Brain White Matter Shape Changes in Amyotrophic Lateral Sclerosis (ALS): A Fractal Dimension Study

Venkateswaran Rajagopalan,Didier Allexandre,Xiao-Feng Wang,Luduan Zhang,Guang H Yue,Erik P Pioro,Zao Liu,Weidong Le

doi:10.1371/journal.pone.0073614

Abstract

Amyotrophic lateral sclerosis (ALS) is a fatal progressive neurodegenerative disorder. Current diagnosis time is about 12-months due to lack of objective methods. Previous brain white matter voxel based morphometry (VBM) studies in ALS reported inconsistent results. Fractal dimension (FD) has successfully been used to quantify brain WM shape complexity in various neurological disorders and aging, but not yet studied in ALS. Therefore, we investigated WM morphometric changes using FD analyses in ALS patients with different clinical phenotypes. We hypothesized that FD would better capture clinical features of the WM morphometry in different ALS phenotypes than VBM analysis. High resolution MRI T1-weighted images were acquired in controls (n = 11), and ALS patients (n = 89). ALS patients were assigned into four subgroups based on their clinical phenotypes.VBM analysis was carried out using SPM8. FD values were estimated for brain WM skeleton, surface and general structure in both controls and ALS patients using our previously published algorithm. No significant VBM WM changes were observed between controls and ALS patients and among the ALS subgroups. In contrast, significant (p<0.05) FD reductions in skeleton and general structure were observed between ALS with dementia and other ALS subgroups. No significant differences in any of the FD measures were observed between control and ALS patients. FD correlated significantly with revised ALS functional rating scale (ALSFRS-R) score a clinical measure of function. Results suggest that brain WM shape complexity is more sensitive to ALS disease process when compared to volumetric VBM analysis and FD changes are dependent on the ALS phenotype. Correlation between FD and clinical measures suggests that FD could potentially serve as a biomarker of ALS pathophysiology, especially after confirmation by longitudinal studies.

Highlights

Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease that affects both upper motor neurons (UMNs) and lower motor neurons (LMNs)
Demographics A total of 100 subject’s data collected as part of our routine clinical scan were assigned into either control or one of the four ALS patient subgroups based on their clinical signs and clinical evaluation of their MRI images: (1) neurological controls, (2) ALS patients with frontotemporal dementia (ALS-FTD), (3) upper motor neuron (UMN)-predominant ALS patients with corticospinal tract (CST) hyperintensity on T2/PD-weighted images (ALSCST+), (4) UMN-predominant ALS patients without CST hyperintensity identified on T2/PD-weighted images (ALSCST–), and (5) ALS-classic patients (ALS-Cl)
The main findings of this study are: 1) No significant difference in Fractal dimension (FD) values was observed between control and any of the ALS subgroups, 2) FD values of WM skeleton and general structure were significantly different between ALS-CST+ and ALS-FTD groups i.e. the FD method was sensitive in identifying differences between ALS subgroups, shown in Figure 1 and Figure 2 (A,B,D &E), 3) In contrast, no significant WM volume changes were observed using voxel based morphometry (VBM) analysis and 4) significant correlation was observed between FD values and ALS functional disability score ALSFRS-R as shown in Figure 4, indicating that FD changes may reflect disease progression; future longitudinal studies may confirm this

Summary

Introduction

Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative disease that affects both upper motor neurons (UMNs) and lower motor neurons (LMNs). Diagnosis of ALS is based on both UMN and LMN degeneration signs. Because no specific test exists to definitively diagnose ALS, diagnosis is based on identifying consistent clinical features and laboratory investigations (e.g., blood tests, EMG, and neuroimaging) to exclude other conditions that mimic ALS [4]. This usually results in significant delay before a definitive diagnosis is made, averaging ,12 months from symptom onset. There has been great interest in identifying biomarkers of ALS, which would allow earlier diagnosis, monitoring disease progression and assessing recognition of efficacy of pharmacotherapies

Objectives

Methods

Results

Discussion

Conclusion