Longitudinal change in structural connectome in Huntington's disease: The image-HD study.

Abstract

Aims To investigate 18 month longitudinal changes in structural connectome in pre-HD and symp-HD, compared with controls. Method Longitudinal analysis of diffusion tensor imaging data was conducted for 28 pre-HD, 25 symp-HD, and 27 controls at baseline and 18 months. Unbiased tensor-based registration pipeline was used to register diffusion tensor data across time and groups. Whole brain probabilistic tractography was performed by seeding 50 million white-matter tracts (streamlines). The SIFT algorithm (spherical-deconvolution informed filtering of tractograms) removed bias in reconstructed tracts. 90x90 structural connectome was generated for each group. Longitudinal change in connectome was analysed using a Network Based Statistics analysis method and a group by time ANOVA model. Results There was a significant within group change in structural connectivity in pre-HD and symp-HD (P < 0.05, FDR Corrected) but not controls. In pre-HD, longitudinal change was observed in two distinct networks: the first connected striatum with motor cortex and the second connected prefrontal cortex with parieto-occipital cortices. In symp-HD, longitudinal change was observed in widespread connexions between frontal and parietal cortices, and striatum and cortex. No distinct networklevel changes were observed in symp-HD. Conclusions Structural connectome analysis can capture network- specific structural disconnectivity. While connectivity change in symp-HD was diffuse overall, the pre-HD group showed distinct networks that changed over time. These connectional pathways offer new avenues to further investigate disease spread pathways in HD.