Cerebral and cerebellar grey matter atrophy in Friedreich ataxia: the IMAGE-FRDA study

Louisa P. Selvadurai, Ian H. Harding, Louise A. Corben, Monique R. Stagnitti, Elsdon Storey, Gary F. Egan, Martin B. Delatycki, Nellie Georgiou-Karistianis

Friedreich ataxia (FRDA) is traditionally associated with neuropathology in the cerebellar dentate nucleus and spinal cord. Growing evidence also suggests involvement of the cerebral and cerebellar cortices, although reports of structural abnormalities remain mixed. This study assessed the structural integrity of cortical grey matter in FRDA, focussing on regions in which pathology may underlie the motor deficits characteristic of this disorder. T1-weighted anatomical magnetic resonance imaging scans were acquired from 31 individuals with FRDA and 37 healthy controls. Cortical thickness (FreeSurfer) and cortical volume (SPM-VBM) were measured in cerebral motor regions-of-interest (primary motor, dorsal and ventral premotor, and supplementary motor areas) alongside unconstrained exploratory analyses of the cerebral and cerebellar cortices. Correlations were assessed between cortical thickness/volume measures and each of disease severity, length of the causative genetic triplet-repeat expansion, and finger-tapping behavioural measures. Individuals with FRDA had significantly reduced cortical thickness, relative to controls, in the premotor and supplementary motor areas. Reduced cortical thickness and/or volume were also observed in the cuneus and precuneus, posterior aspects of the medial and lateral prefrontal cortices, insula, temporal poles, and cerebellar lobules V, VI, and VII. Measures of clinical severity, genetic abnormality, and motor dysfunction correlated with volume loss in the lateral cerebellar hemispheres. These results suggest that atrophy preferentially affects premotor relative to primary areas of the cortical motor system, and also extends to a range of non-motor brain regions. Furthermore, cortical thickness and cortical volume findings were largely divergent, suggesting that each is sensitive to different aspects of neuropathology in FRDA. Overall, this study supports a disease model involving neural aberrations within the cerebral and cerebellar cortices, beyond those traditionally associated with this disorder.