E05 6-,9- And 15-Month Change in Cortical Thickness and Region-Of-Interest Volume And Diffusion Metrics in Huntington's Disease: Informing Trial Design

Abstract

Background Major efforts are being invested in the development of disease-modifying compounds for Huntington’s disease (HD). Testing the efficacy of such compounds in clinical trials is dependent on the validation of biomarkers capable of detecting disease progression over short time frames. Aim To evaluate a range of neuroimaging, clinical and cognitive markers of HD progression over 6-, 9- and 15-month intervals. To present effect sizes (ES) for each outcome which can be used to estimate sample sizes, and hence appropriately power, future disease-modifying trials. Method 40 controls and 61 stage 1 HD patients underwent 3T MRI and standard clinical and cognitive assessments at baseline, 6- and 15-months. Neuroimaging analysis included longitudinal change in lobar cortical thickness, striatal-, corpus callosal- and global-volume and diffusion metrics. ES were calculated for the difference in change between HD and controls. Results Volumetric neuroimaging measures such as caudate, white-matter and whole-brain atrophy produced consistently large ES over 6-, 9- and 15-month intervals. Diffusion metrics were slightly less sensitive, although caudate diffusivity metrics performed strongly over 15 months. Cortical thickness metrics lacked sensitivity to change over all lobes and all time intervals. Clinical and cognitive outcomes tended to lack sensitivity, particularly over 6- and 9-months, with wide confidence intervals around ES estimates. Conclusion Volumetric measures such as caudate atrophy may have utility as short-term readouts in proof-of-concept studies over 6 months and as secondary outcomes in larger efficacy studies over longer intervals. Specific diffusion metrics have potential utility as efficacy readouts over 9- and 15-months and may be of particular use if the mechanism of action of the therapy is relevant to white-matter integrity. Lobar cortical thickness readouts do not appear to provide sensitive markers of progression over these intervals.