Objective: In previous imaging studies, we have that patients with Parkinson’s disease (PD) express a distinct brain network known as the Parkinson’s disease-related pattern (PDRP). Here, we used resting state fMRI (rs-fMRI) and graph analysis to show a stereotyped change in the wiring of the PDRP network in individual PD patients.

Background: The PDRP network, originally characterized using metabolic PET, has subsequently been identified using resting state fMRI (rs-fMRI) [1,2]. While graph theoretic methods have previously being used to evaluate network organization in PD [3,4], it is unclear whether the observed changes reflect a pathological connectivity response that can be discerned in individual patients. To address this issue, we used a new approach to analyze rs-fMRI data from two independent samples of PD and healthy control (HC) subjects.

Method: We studied a cohort of 20 PD and 20 age- and gender-matched HC subjects who underwent rs-fMRI at North Shore University Hospital (NSUH) [table1]. For validation, we analyzed baseline rs-fMRI scans from 102 PD and 18 HC [table1] who were studied as part of Parkinson’s Progression Markers Initiative (PPMI; https://www.ppmi-info.org/). We used the previously reported rs-fMRI PDRP (fPDRP [1]) composed of 39 nodes [table2] [5], to measure fPDRP assortativity, a descriptor of the pattern (intercorrelation) of nodal connectivity within a network [6], in the individual scan data. Conventional graph metrics (degree centrality, clustering coefficient, characteristic path length, and small-worldness) were likewise evaluated at the subject level. fPDRP assortativity was measured in the individual scans and group differences were evaluated across graph thresholds using the general linear model.

Results: In the fPDRP network [figure1A], PD exhibited significant increases in assortativity (P<0.005) compared to HC scanned at NSUH [figure1B]. The findings were confirmed in the larger PPMI testing set. As with NSUH dataset, fPDRP assortativity was significantly elevated in these patients (P<0.001) compared to HC [figure1C]. Group differences were inconsistent for the other graph  metrics.

Conclusion: Increased assortativity is consistent feature of the PD-related network. Information flow through high assortativity networks tends to be unstable and inefficient. The performance of pathological networks such as PDRP may be improved by therapeutic rewiring strategies.

References: 1. Vo, A. (2017), ‘Parkinson’s disease-related network topographies characterized with resting state functional MRI’, Hum Brain Mapp, vol. 38, pp. 617–630. 2. Schindlbeck, KA. (2018), ‘Network imaging biomarkers: insights and clinical applications in Parkinson’s disease’, Lancet Neurol. vol. 17, pp. 629–640. 3. Niethammer, M. (2018), ‘Gene therapy reduces Parkinson’s disease symptoms by reorganizing functional brain connectivity’, Sci Transl Med, vol. 10, no. 469. 4. Schindlbeck, KA. (2020), ‘LRRK2 and GBA Variants Exert Distinct Influences on Parkinson’s Disease-Specific Metabolic Networks’, Cerebral Cortex, vol. 30, pp. 2867–2878. 5. Tzourio-Mazoyer, N. (2002), ‘Automated anatomical labeling of activations in SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject brain’, Neuroimage, vol. 15, pp. 273–289. 6. Noldus, R. (2015), ‘Assortativity in complex networks’, J. Complex Networks. vol. 3, pp. 507–542.

« Back to MDS Virtual Congress 2021

MDS Abstracts - https://www.mdsabstracts.org/abstract/resting-state-functional-mri-network-assortativity-is-increased-in-parkinsons-disease/