Association of in vivo retention of [18f] flortaucipir pet with tau neuropathology in corresponding brain regions

Abstract

[¹⁸F]Flortaucipir is an FDA-approved tau-PET tracer that is increasingly utilized in clinical settings for the diagnosis of Alzheimer’s disease. Still, a large-scale comparison of the in vivo PET uptake to quantitative post-mortem tau pathology and to other co-pathologies is lacking. Here, we examined the correlation between in vivo [¹⁸F]flortaucipir PET uptake and quantitative post-mortem tau pathology in corresponding brain regions from the AVID A16 end-of-life study (n = 63). All participants underwent [¹⁸F]flortaucipir PET scans prior to death, followed by a detailed post-mortem neuropathological examination using AT8 (tau) immunohistochemistry. Correlations between [¹⁸F]flortaucipir standardized uptake value ratios (SUVRs) and AT8 immunohistochemistry were assessed across 18 regions-of-interest (ROIs). To assess [¹⁸F]flortaucipir specificity and level of detection for tau pathology, correlations between [¹⁸F]flortaucipir SUVR and neuritic plaque score and TDP-43 stage were also computed and retention was further assessed in individuals with possible primary age-related tauopathy (PART), defined as Thal phase ≤ 2 and Braak stage I–IV. We found modest-to-strong correlations between in vivo [¹⁸F]flortaucipir SUVR and post-mortem tau pathology density in corresponding brain regions in all neocortical regions analyzed (rho-range = 0.61–0.79, p < 0.0001 for all). The detection threshold of [¹⁸F]flortaucipir PET was determined to be 0.85% of surface area affected by tau pathology in a temporal meta-ROI, and 0.15% in a larger cortical meta-ROI. No significant associations were found between [¹⁸F]flortaucipir SUVRs and post-mortem tau pathology in individuals with possible PART. Further, there was no correlation observed between [¹⁸F]flortaucipir and level of amyloid-β neuritic plaque load (rho-range =  – 0.16–0.12; p = 0.48–0.61) or TDP-43 stage (rho-range =  – 0.10 to  – 0.30; p = 0.18–0.65). In conclusion, our in vivo vs post-mortem study shows that the in vivo [¹⁸F]flortaucipir PET signal primarily reflects tau pathology, also at relatively low densities of tau proteinopathy, and does not bind substantially to tau neurites in neuritic plaques or in individuals with PART.