Discordant low amyloid‐β PET and high neocortical tau PET retention

Abstract

AbstractBackgroundExtracellular amyloid‐β (Aβ) plaques and intracellular tau neurofibrillary tangles characterize Alzheimer’s disease (AD). Combined use of Aβ and tau PET biomarkers demonstrates that high neocortical tau is rarely observed in the absence of high Aβ. We describe the prevalence of participants with high neocortical tau PET retention (visually and quantitatively) with a correspondingly low Aβ PET result and characterize these participants.MethodAIBL participants who had both an Aβ and tau PET scan were included. Centiloid <25 defined a low (negative) Aβ result. Tau PET SUVR (standardized uptake value ratio) was estimated in three composite regions of interest (ROI): mesial temporal (Me), temporoparietal (Te) and rest of neocortex (R). The 99%ile of the Aβ‐ve cognitively unimpaired participants determined the threshold for each ROI.ResultFour hundred and sixty‐six participants were included. Of all Aβ‐ participants (n=287), 12 were above the threshold in Te and R ROIs. Four participants (1.4% of all Aβ‐) had high neocortical tau tracer retention, quantitatively and visually, in a pattern consistent with AD (Braak stage V‐VI) (Figure 1). Despite the low Aβ PET, one participant had high plasma Aβ, while another had low CSF Aβ1‐42, high CSF pTau and tTau and an 18F‐FDG PET pattern consistent with AD. High plasma pTau‐217‐plus was observed for three of these four participants (result unavailable for one participant).ConclusionWe have identified four participants with low Aβ PET and high neocortical tau PET retention. The alternative biomarker results in these cases are suggestive of a diagnosis of AD, despite the low Aβ PET result. We propose that in these instances, the Aβ PET ligand may not be detecting Aβ, maybe due to a different conformation of the aggregates. While infrequent, this finding has been reported using 11C‐PiB. Overall, the prevalence of this discordance was <2% in the sampled cohort, suggesting it is an uncommon occurrence. However, this study is limited by the lack of alternative Aβ biomarker results. Studies in larger cohorts are necessary to validate these findings, while genetic testing and postmortem correlation will enhance our understanding of the potential explanation for this finding.