Asymmetry and vulnerability in amyotrophic lateral sclerosis

Abstract

Objectives: Amyotrophic lateral sclerosis (ALS) is a progressive, fatal condition in which both upper (UMN) and lower motor neurons (LMN) undergo irreversible degeneration. Although the cause remains unknown and the disease is phenotypically diverse, certain functions and muscle groups are particularly vulnerable to early weakness, including the thenar eminence and foot dorsiflexion. It is also well established that disease onset and spread are typically asymmetric. Although this may reflect an additional source of vulnerability, the factors underlying this asymmetry are unknown. Furthermore, separate analysis of the asymmetry of UMN and LMN signs has not been performed, nor has this concept been applied to non-limb features of ALS such as bulbar involvement.The aim of this project was to identify and characterise novel sources of vulnerability in a large cohort of ALS subjects, with particular focus on asymmetry, clinical phenotype and the effect of limb dominance on both onset and spread. Multiple modalities were employed, including structured questioning, direct clinical examination and scoring, and advanced magnetic resonance imaging (MRI) analysis. As a secondary objective, the burden of clinical LMN and UMN involvement was also correlated with patient survival across the study period.Methods: 161 consecutive subjects with ALS were recruited from two tertiary centres. The effect of limb dominance on the onset and spread of weakness was assessed using a standardised structured questionnaire, followed by non-parametric statistical analysis. The clinical severity of UMN and LMN signs in each limb was then determined using a published scoring system. These scores were used to quantify UMN and LMN dysfunction in dominant and non-dominant limbs. The differential involvement of dysphagia and dysarthria in bulbar ALS was also investigated in a subgroup (n = 39).To further investigate the laterality of central nervous system (CNS) changes, a novel voxel-based morphometry (VBM) MRI protocol was used, previously utilised to define normal cortical gray matter (GM) asymmetries in healthy subjects. In the current study, it was applied to a cohort of right-handed healthy controls (n = 17), and ALS subjects with first weakness in either a right-sided or left-sided limb (n = 15 each). Between-group voxelwise comparisons of GM density were performed. Subsequently, within-group comparisons were performed to assess areas of GM asymmetry between the two hemispheres in both healthy controls and ALS subjects.Finally, the variables affecting survival of the original cohort (n = 161) were investigated, in particular the effects of UMN and LMN scores. Subjects were followed to a censorship date, after which univariate analysis was performed to screen potential predictor variables of survival to non-invasive ventilation (NIV) or death. Potentially useful predictor variables were then used in a Cox regression model.Results: Onset of weakness was more likely in the dominant upper limb (p = 0.02), but not in the dominant lower limb (p = 0.78). Furthermore, there was a significant effect of limb dominance on spread of weakness beyond the initial limb. For example, in subjects with initial weakness in a non-dominant limb, spread was more likely to be to the other ipsilateral limb (p = 0.008), suggesting an important role for central (UMN) factors in driving disease spread.Both UMN and LMN scores were maximal in the limb of onset, suggesting focality of onset. UMN signs spread to other limbs relatively early. A significant effect of dominance was also identified, specifically that the distribution of UMN signs in the upper limbs was affected by whether weakness had first occurred on the dominant or non-dominant side (p = 0.03). There was no significant effect of limb dominance on the distribution of LMN signs.Voxelwise MRI analysis revealed multifocal clusters of reduced GM density in ALS subjects compared with controls, incorporating both motor and non-motor areas. Normal regions of cortical asymmetry in right-handed healthy controls were also identified, including leftward GM asymmetry (p ≤ 0.01) corresponding to the dominant hand representation area in the motor cortex. Right-handed subjects with ALS showed complete loss of GM asymmetry in this area, irrespective of whether weakness had first occurred in a dominant or non-dominant limb (p < 0.01). However, asymmetric atrophy of the left somatosensory cortex and temporal gyri was only observed in ALS subjects with right-sided (dominant) onset of weakness.Predictor variables associated with reduced survival included older age, bulbar and respiratory involvement, and shorter diagnostic delay (all p < 0.05). Whole-body LMN scores were strongly associated with survival to NIV or death, whereas UMN scores were poorly associated with survival.Conclusions: Both the clinical and imaging findings support limb dominance as a significant factor underlying both onset and spread of ALS, with central (UMN) processes playing an important role in disease asymmetry. This effect of limb dominance on the presentation of ALS may reflect underlying developmental CNS vulnerabilities, which become exposed by the disease process. However, ultimately LMN factors more closely correlate with survival.