Longitudinal tau-PET accumulation and downstream effects in amyloid-beta negative cognitively unimpaired individuals

BACKGROUND: Previous studies, including ours (Wuestefeld et al., Brain, 2023), showed tau positron emission tomography (PET) uptake in medial temporal and neocortical regions of amyloid-beta (Aβ)-negative cognitively unimpaired (CU) individuals, associated with neurodegeneration and worse memory performance. To further understand if tau-PET uptake is clinically relevant in this population, we characterized Aβ-negative CU individuals with higher longitudinal tau-PET accumulation and its associations with atrophy and cognitive decline. METHOD: We included 333 CU BioFINDER-2 participants, negative for both global [18F]flutemetamol Aβ-PET and cerebrospinal fluid (CSF) Aβ42/Aβ40 (age=63.8, 55% female, 2.28±1.36 years follow-up). Using linear mixed-effects models, we calculated the rate of change (ROC) in [18F]RO948 tau-PET SUVR in composite regions of interest (ROI), recapitulating the Braak stages. Individuals were classified as "tau-accumulators" vs. "non-accumulators" based on the mean+standard deviation of Braak I-IV tau-PET ROC in young controls (20-40 years, n = 29; Figure 1A). Tau-PET ROC and thickness of the entorhinal cortex, Brodmann area (BA)35 and neocortical AD-regions (precuneus/posterior cingulate, lateral occipital, superior frontal) were extracted. Group differences and associations with CSF AD biomarkers, delayed word-list recall, and modified Preclinical Alzheimer Cognitive Composite were tested. RESULT: 16% of individuals were identified as tau-accumulators. They were older, had lower levels of Aβ as indicated by a higher CSF Aβ42/40 ratio (no difference on Aβ-PET) and numerically higher CSF MTBR-tau243/Aβ40 (not significant, note missingness; Figure 1B+C). Tau-accumulators showed significantly greater tau-PET at baseline and accumulation in regions not included in Braak I-IV (all p <0.001). Additionally, we observed more atrophy in BA35 (Figure 2A, not after age-adjustment) and numerically worse baseline cognitive performance, but no difference in decline over time (Figure 2B). In the whole sample, Braak I-IV tau-PET accumulation was significantly associated with increased age, more BA35 atrophy and at trend-level with higher levels of CSF Aβ42/40 and CSF MTBR-tau243/Aβ40 (Figure 3). CONCLUSION: A subgroup of Aβ- CU individuals show longitudinal tau accumulation across the neocortex. Preliminary results suggest an association with a tau-specific CSF biomarker and focal atrophy, which did not translate to changes in cognition. We will complement our analyses with a data-driven approach of classifying tau-accumulators.