Evaluation of neuroimaging biomarkers in amyotrophic lateral sclerosis

Abstract

Amyotrophic lateral sclerosis (ALS) is a progressive neurodegenerative condition characterised by relentless upper and lower motor neuron degeneration. The clinical spectrum, symptoms onset, genetic vulnerability and cognitive profile of ALS is hugely heterogeneous making early diagnosis, accurate prognosis and disease monitoring particularly challenging. There is no cure available and life expectancy is limited to 2-3 years. Objective, accurate and validated biomarkers are urgently needed for diagnostic applications, disease monitoring and as prognostic indicators.

The aim of this PhD thesis is to evaluate the role of magnetic resonance imaging (MRI) as a potential biomarker of ALS.

First, a comprehensive systematic literature review was conducted to explore the methods, design and pitfalls of existing longitudinal imaging studies across the spectrum of neurodegenerative conditions including Alzheimer's disease, amyotrophic lateral sclerosis, fronto-temporal dementia, Huntington disease, multiple sclerosis, Parkinson's disease, ataxia, HIV, alcohol dependence and healthy ageing.

Subsequently, ALS-associated structural brain changes were evaluated crosssectionally and longitudinally. The results were used in diagnostic and prognostic models to assess the biomarker potential of MRI metrics. The multimodal analyses relied on complementary grey and white matter measures based on T1-weighted and diffusion-weighted images. The methods included voxel-based morphometry, vertex-based cortical thickness analyses and track-based analyses of fractional anisotropy, mean -, radial - and axial diffusivity indices.

The most prominent imaging features identified in this ALS study are the progressive degeneration of the corpus callosum, corticospinal tracts, corticobulbar fibres and the precentral gyri.

The comprehensive analyses of longitudinal MRI data highlight progressive grey matter alterations and suggest that grey matter metrics are better suited for monitoring purposes than white matter indices.

Based on these imaging patterns, an automated classification protocol to distinguish blinded MR data sets of ALS patients and healthy controls with good accuracy and sensitivity was developed.

Finally, the prognostic value of MRI was demonstrated predicting 18-month survival based on MR data sets alone and in combination with clinical data. The supplementary prognostic value of neuroimaging measures in addition to more-established clinical and demographic prognostic factors was shown.

In conclusion, it was demonstrated that MRI can be used as a diagnostic or prognostic biomarker and it was shown that monitoring markers should focus on grey matter degeneration.