Subcortical tau‐accumulation predicts neuronal dysfunction in the cortex based on functional connectivity in 4R‐tauopathies

Abstract Background 4‐repeat (4R) tauopathies are neurodegenerative diseases characterized by cerebral accumulation of 4R tau pathology. The most prominent 4R‐tauopathies are progressive‐supranuclear‐palsy (PSP) and corticobasal‐syndrome (CBS) characterized by tau accumulation in subcortical nuclei as well as cortical neuronal dysfunction, as shown by PET‐assessed hypoperfusion and glucose hypometabolism. Yet, there is a spatial mismatch between subcortical tau deposition patterns and cortical neuronal dysfunction and it is unclear how these two pathological brain changes are interrelated. Here, we hypothesized that subcortical tau pathology induces diaschisis‐like neuronal dysfunction in functionally connected cortical regions. Method We included 47 patients with clinically diagnosed PSP or CBS who underwent structural MRI and 18 F‐PI‐2620 tau‐PET. PI‐2620 PET was recorded using a dynamic one‐shot, two‐stop acquisition protocol, to determine an early 0.5‐2.5min post‐tracer‐injection perfusion window for assessing cortical neuroinjury in 200 cortical ROIs of the Schaefer atlas, as well as a 20‐40min post‐tracer‐injection window to determine 4R‐tau load in 32 subcortical ROIs of the TIAN atlas. We determined tau epicenters as 10% of subcortical ROIs with highest tau‐PET, and assessed the connectivity of tau epicenters to cortical ROIs using an age‐matched 3T resting‐state fMRI template derived from 69 healthy elderly. Using linear regression, we assessed whether i) higher subcortical tau‐PET was associated with overall reduced cortical perfusion and ii) whether cortical hypoperfusion was observed preferentially in regions closely connected to subcortical tau epicenters. Result As hypothesized, higher subcortical tau‐PET was associated with lower cortical perfusion (R=‐0,37, p‐value: <0,011, Fig.1). Using group‐average tau‐PET and perfusion‐PET, we found that the seed‐based connectivity pattern of subcortical tau epicenters predicted cortical perfusion patterns, where cortical regions that were more closely connected to the tau epicenter showed stronger hypoperfusion (R=‐0,16, p‐value: <0,023, Fig.2A). This association was also observed on the subject level, as indicated by overall negative b‐values of the association between tau epicenter connectivity and cortical perfusion (one‐sample t‐test: t‐value: ‐3,45, p‐value: <0,001, Fig.3). Conclusion In 4R‐tauopathies subcortical tau‐accumulation is associated with remote neuronal dysfunction in functionally connected cortical regions. This suggests that subcortical tau pathology may induce diaschisis‐like cortical dysfunction, which may contribute to clinical disease manifestation and clinical heterogeneity.