Disrupted Rich-Club Organization of Brain Structural Networks in Parkinson’s Disease

Liu Tiantian,Yan,Ai Jing,Chen Duanduan,Beijing Institute of Technology,Fang Boyan,Yan Tianyi

DOI: https://doi.org/10.1007/s00429-021-02319-3

2021-01-01

Brain Structure and Function

Abstract:Parkinson’s disease (PD) can be considered as the dysfunction in segregation and integration of large-scale structural networks in the late stage of disease progression. However, the altered patterns in the early stage have not been extensively investigated, especially the altered structural rich-club patterns, which is proved powerful to detect the altered patterns of structural networks in Alzheimer’s disease and schizophrenia. To this end, we investigated the rich-club organization of the structural networks derived from diffusion tensor imaging (DTI) data in the early stage of PD, and further investigated the relationship between rich-club organization and clinicopathological measures, including motor and non-motor scales and cerebrospinal fluid (CSF) biomarkers. Two datasets were included for validation in this study. The first one included 41 healthy controls (HC) and 64 PD patients from Parkinson’s Disease Progression Marker Initiative (PPMI) dataset, and the second one included 24 HC and 26 PD patients. Results revealed that PD patients in early stage had disrupted rich-club organization, with abnormal connectivity strength between peripheral regions (two-sample t-test between PD and HC: p < 0.001), whereas connectivity strength between rich-club regions remained relatively stable (two-sample t-test between PD and HC: p = 0.108). The classification accuracies on three types of connections were 59.93%, 73.96% and 77.44% for rich-club, feeder and local connections. Furthermore, abnormal local and feeder connections showed significant correlation with poor clinical scales and CSF biomarkers. In summary, a selective disruption of non-rich-club connections here could be regarded as a potential marker in the early diagnosis of PD.