Structural and diffusion magnetic resonance imaging in mitochondrial disease : validation, optimisation, and implementation

Abstract

Mitochondrial DNA mutations lead to a group of diseases that are both genetically and phenotypically heterogeneous. The m.3243A>G point mutation is one of the most common and is also present in around 80% of patients with mitochondrial encephalomyopathy, lactic acidosis and stroke-like episodes (MELAS), which is a phenotype characterised by symptoms that affect multiple systems and are either progressive or recessive progressive in nature. The aim of this thesis was to develop analysis methodologies to quantitatively characterise the structural properties of the brains of controls, carriers, and MELAS patients using non-invasive MRI techniques to then determine I developed robust segmentation protocols from first principles for the global segmentation using FSL 5.0, and also for the regional segmentation using FreeSurfer 5.4. Global segmentation results shows that total intracranial volume may provide a surrogate marker for susceptibility to the onset of stroke-like episodes. Grey and white matter volumes showed significant difference in brain composition between controls and carriers despite the absence of overall brain volume differences. Regional segmentation highlighted that MELAS patients showed an almost global thinning of the cortex compared to both controls and carrier, but all carriers of the m.3243A>G point mutation showed significant thinning of the temporal pole and entorhinal cortex irrespective of the symptoms presented. The TBSS analysis repeated this finding of a global difference in white matter fractional anisotropy between carriers and MELAS patients, indicating that although stroke-like episodes lead to localised lesions, the degenerative effects are felt throughout the brain. The ultra-high-resolution MRI of post mortem sections highlighted the potential for the use of this technique in the future to allow the co-registration of MRI and neuropathological imaging despite the condition of the brain sections being suboptimal due to mounting in hydrogel. This study provides a protocol for the systematic, quantitative characterisation, along with initial data capture and analysis, of brain volumes in people harbouring the m.3243A>G point mutation. This can be used as a starting point to more effectively investigate much larger populations of people with mitochondrial disease.