• DTI
• functional connectivity
• graph theory
• non-AD tauopathy
• PET/MRI
• probabilistic tractography
• PSP
• resting-state fMRI
• structural connectivitiy
• tau
• tau-PET
• tau-PET tracer
• [18F]PI-2620

Introduction
Progressive supranuclear palsy (PSP) is a primary 4-repeat (4R) tauopathy clinically characterized by postural instability with falls, akinesia, ocular motor dysfunction and cognitive decline. The clinical symptoms of PSP have strong overlaps with other neurodegenerative disorders and the diagnosis of definite PSP is only post mortem by establishing region-specific tau aggregates predominantly in the brainstem, basal ganglia and cerebellar nuclei. The novel second-generation tau-PET tracer [18F]PI-2620 showed a high binding affinity for aggregated 3/4R tau in Alzheimer’s disease (AD), as well as high affinity to recombinant 4R tau fibrils in non-AD tauopathies. [18F]PI-2620 imaging also proved to aid in diagnosing and differentiating patients with suspected PSP, potentially facilitating a more reliable diagnosis of PSP. Besides, there is gathering evidence supporting the concept of disrupted and reorganized structural and functional connectivities in PSP that might be influenced by tau pathology. Therefore, the present PET/MRI study with novel tau-PET tracer, we aimed to assess the structural and connectivity network properties in PSP patients and their relation to tau burden.
Material and methods
Ten PSP patients (age 72±7 years, 6 female) underwent 60 minutes of dynamic PET imaging following 300 MBq bolus injection of [18F]PI-2620 on an integrated 3T PET/MRI system. The binding potential (BP) was estimated using non-invasive pharmacokinetic modeling (MRTM2 in PMOD). Diffusion tensor imaging (DTI) and resting-state fMRI were acquired simultaneously and compared to the measurements of ten healthy controls (age 65±8 years, 4 female). Two networks: the whole-brain and deep gray matter (DGM) networks have ween generated. For the whole-brain network, we used default atlas-based (a combination of the Harvard-Oxford atlas and the AAL atlas) definition of 132 cortical and subcortical region of interests (ROIs), while for the DGM network, a custom atlas was generated consisting of thirteen PSP-target regions: putamen, globus pallidum externus (GPe), substantia nigra (SN), subthalamic nucleus (STN), periaqueductal gray matter, red nucleus (RN) and dentate nucleus (DN). A graph theory approach and functional network connectivity (FNC) analysis were applied to investigate PSP-related structural and effective functional connectivities, derived respectfully, from probabilistic tractography (network mode) and ROI-to-ROI functional connectivity (RRC) analysis. A possible association between tau load and various network attributes were studied. For voxel-wise analysis, between-subject effect and regression model were designed using the General Linear Model (GLM) and a false discovery rate (FDR) for multiple testing correction of pFDR < 0.05 was applied. In addition, Student's t-test and Spearman's rank correlation coefficient were used, respectively, for between-group differences of the global network attributes and their association with BP.
Results
For whole-brain structural connectivity, at the global network level, the number of triangles, density, and the spatial distance was lower, while assortativity degree and vulnerability were higher in PSP patients compared to controls (p<0.05). At the vertex level, transitivity, edge asymmetry, average nearest neighbor strength, “hubness” scores, vulnerability, and eccentricity showed variable group differences (pFDR<0.05). Positive relationships were found between BP in the right cerebellar crus and clustering coefficient (r=0.74, p<0.05); BP in the left rolandic operculum and the global efficiency (r=0.65, p<0.05). Inverse relationships were observed between BP in the right precentral, bilateral frontal regions, caudate and subthalamic nuclei and the network vulnerability (r>0.60, p<0.05). The “hubness” scores demonstrated a negative association with BP in the left cerebellar crus (r=-0.76, p<0.05).
For whole-brain functional connectivity, 17292 functional connections among 132 ROIs were analyzed. The between-subject effect depicted three clusters, showing increased connectivity between frontal-orbital cortex and cerebellar vermis and decreased connectivity between putamen and mid-temporal gyrus in PSP patients compared to controls (pFDR < 0.05). Regression analysis showed a negative relationship between BP in the bilateral DN and cerebellar-putaminal/cerebellar-pallidal connectivity (pFDR < 0.05).
For deep grey network structural connectivity, at the vertex level, local efficiency was lower in SN/RN (left) and bilateral DN; average path length was decreased in STN (left) and putamen (right), while increased in ST/RN (left) and GPe (right); vertex strength was higher in the bilateral ST; nodal vulnerability was increased in STN/DN (left) and putamen (right), while decreased in all other regions. The BP showed an inverse relationship with nodal assortativity for bilateral putamen/RN/STN and with vulnerability for STN (left).
For deep grey network functional connectivity, at the global network level, the global and local efficiencies, clustering coefficient, and betweenness centrality were significantly lower, while average path length and degree were higher in PSP compared to controls. At the vertex level, global efficiency was lower for putamen/ST (left) and STN/RN (right); average path length and degree were higher for putamen/ST (left). The BP showed a strong relationship (r=-0.7, p<0.05) with local efficiency and clustering coefficient for all nodes; additionally, the STN (left) showed an association with global efficiency, degree, average path length, and betweenness centrality.

Conclusion
Graph theory and FNC analysis of the structural and functional connectivities depicted various divergences both for whole-brain and DGM network properties, indicating lower network integration and higher segregation in PSP patients compared to controls. Aberrant frontal-cerebellar functional connectivity was presented in PSP patients that seem to be influenced by tau-pathology in the cerebellar DN. The PSP-related tau pathology appears to be associated with different measures of network structures. Cerebellum and cerebellar crus might impede the smoothness of the information passage in association with increased tau load. These results support the network degradation/reorganization concept in response to tau pathology and motivate future investigations in a larger PSP patient cohort.