MRI as a biomarker in the presymptomatic, diagnostic and disease progression phases of the ALS-FTD spectrum

This thesis investigated whether MRI can serve as a biomarker across the presymptomatic, diagnostic, and disease progression phases of the ALS-FTD spectrum. Each phase poses distinct measurement challenges, but all share a lack of a disease-specific biomarker. Current tools such as EMG, blood biomarkers, questionnaires, and cognitive screens suffer from invasive acquisition, interrater variability, low specificity, or symptom masking. Quantitative MRI is non-invasive, rater-independent, and unaffected by symptom masking, and because MRI is already used clinically to exclude mimics, repurposing the same data adds little cost.
Presymptomatic phase. In C9orf72 expansion carriers without symptoms (AFM C9+), brain MRSI revealed neuronal and glutamatergic dysfunction, altered glycogen metabolism, and increased membrane turnover, with motor cortex sparing. ALS C9+ patients showed the same pattern but more severe, widespread, and motor-cortex-involving, while ALS C9- patients showed glutamatergic dysfunction, energy metabolism disturbance, and reduced phospholipids. Lower tATP/PCr and mI+Gly/tCr emerged as candidate markers of phenoconversion. These findings support testing Riluzole and energy-metabolism-targeted agents in presymptomatic carriers. A longitudinal structural MRI study further quantified individual presymptomatic cortical atrophy up to six years before symptom onset, and predicted that 27.9% of non-converted AFM C9+ carriers will show phenoconversion-associated atrophy by age 80, closely matching recent penetrance estimates.
Diagnostic phase. Quantitative spinal cord MRI distinguished LMN-ALS from its closest clinical mimic, MMN, and correlated with detailed neurological examination of 24 muscle groups. In vivo ventral horn degeneration was demonstrated, bringing spinal cord MRI one step closer to becoming a disease-specific biomarker that could reduce diagnostic delay. Using brain MRI in a large ALS cohort, two clinically recognized phenotypes (pure motor; motor with frontotemporal involvement) were confirmed, and a novel cingulate-parietal-temporal-cerebellar (CPT) phenotype was identified, likely missed by ECAS because that screen targets frontotemporal impairment.
Disease progression phase. Focused motor cortex MRI detected UMN degeneration a median of 4.4 months before clinical UMN signs appeared, and was more sensitive than neurological examination in 4 of 5 body regions. For clinical trial endpoints, 41 patients scanned with MRI would match the statistical power of 167 patients assessed by neurological examination, a 4.1-fold reduction with major cost implications. Longitudinal multimodal MRI also revealed distinct structural and network correlates of each cognitive and behavioural domain (including a separate social cognition pattern) and detected significant decline where ECAS and ALSFTD-Q did not, indicating MRI captures cognitive-behavioural progression more sensitively than repeated screens.
Future perspectives. Follow-up work will study Riluzole effects using MRSI, apply deuterium-labelled glucose MRSI to energy metabolism, extend individual atrophy prediction to all neuromuscular clinic referrals, replicate spinal cord findings earlier in the disease course, and embed quantitative brain MRI prospectively and retrospectively in clinical trials. A proposed synergy framework would share MRI methodology across Dutch neurodegenerative expertise centres to identify common pathways and enable parallelized multi-disease trial platforms, accelerating the path to effective therapies for patients across the ALS-FTD spectrum and beyond.