Defining and characterizing neocortical tau resistance in preclinical Alzheimer’s Disease

Abstract

AbstractBackgroundTopographical staging of tauopathy in preclinical Alzheimer’s disease (AD) posits early deposition in the entorhinal region (medial temporal lobe; MTL) followed by spread to adjacent neocortical (NEO) regions. Predictive models support a cascade of tau deposition as baseline MTL tau robustly predicts NEO tau accumulation. In clinically normal (CN) Ab+, elevated tau in MTL and NEO is associated with accelerated progression to dementia relative to those with only elevated MTL tau. The frequency and profile of neocortical tau resistance (defined as MTL+/NEO‐) remain unclear.MethodWe defined tau resistance using two longitudinal datasets. To compare directly across MTL (entorhinal/amygdala/parahippocampal) and NEO (inferior‐temporal/fusiform/middle‐temporal/inferior‐parietal) composites, we averaged regional SUVrs within each composite then normed them to independent samples of younger CNAb‐ adults. We selected CN participants from ADNI and HABS with ≥2 time‐points of Flortaucipir‐PET (nHABS = 200; nADNI = 149; cerebellar‐greyreference). We extracted slopes from a linear mixed model covarying for age, APOEe4 and Abstatus. We compared baseline MTLnormed values with slopes in a NEOnormed. We defined tau resistance as those with elevated baseline MTLnormed tau‐PET but no change in NEOnormed tau‐PET over time. We used non‐parametric group comparisons to determine differences in a range of variables (Table1).ResultWe found a poor to moderate correlation between baseline MTLnormed values and adjusted NEOnormed slopes (rHABS = 0.12; rADNI = 0.36; Fig.1), with a non‐linear association at higher levels of MTLnormed. In HABS, tau resistance was identified in 13% (n = 10) of those with elevated MTLnormed. There was a trend for Black/African American individuals to be less tau resistant (Table.1). In ADNI, 26% were identified (n = 11), there was a trend toward less APOEe4 carriers in the resistant group. Although the ranges of change were different across the two cohorts, adjusted MTLnormed and NEOnormed slopes were similarly strongly correlated (rHABS = 0.69; rADNI = 0.74; Fig.2).ConclusionOur preliminary findings suggest that resistance to neocortical tau in CN exists in up to a quarter of those with elevated MTL. Given the complexities with determining cross‐sectional and longitudinal tau‐PET thresholds, our approach to identifying MTL+ or NEO+ may be too conservative. Future work will validate tau resistance via cognitive decline and clinical progression.