Alzheimer's Imaging Consortium.

Background: Tau-PET imaging allows in-vivo detection of neurofibrillary tangles. One tau-PET tracer (i.e., [18F]flortaucipir) has received FDA-approval for clinical use, and multiple other tau-PET tracers have been implemented into clinical trials for participant selection and/or as a primary or secondary outcome measure. To optimize future use of tau-PET, it is essential to understand how demographic, clinical and genetic factors affect tau-PET-positivity rates.

Method: This large-scale multi-center study includes 9713 participants from 35 cohorts worldwide who underwent tau-PET with [18F]flortaucipir (n = 6420), [18F]RO948 (n = 1999), [18F]MK6240 (n = 878) or [18F]PI2620 (n = 416) (Table-1). We analyzed individual-level tau-PET SUVR data using a cerebellar reference region that were processed either centrally (n = 3855) or by each cohort (n = 5858). We computed cohort-specific SUVR thresholds based on the mean + 2 standard deviations in a temporal meta-region of amyloid-negative cognitively normal (CN) individuals aged >50. Logistic generalized estimating equations were used to estimate tau-PET-positivity probabilities, using an exchangeable correlation structure to account for within-cohort correlations. Analyses were performed with (interactions between) age, amyloid-status, and APOE-e4 carriership as independent variables, stratified for syndrome diagnosis.

Result: The study included 5962 CN participants (7.5% tau-PET-positive), 1683 participants with mild cognitive impairment (MCI, 33.8% tau-PET-positive) and 2068 participants with a clinical diagnosis of dementia (62.1% tau-PET-positive) (Figure-1). From age 60 to 80 years, the estimated prevalence of tau-PET-positivity increased from 1.2% [95% CI: 0.9%-1.5%] to 3.7% [2.3%-5.1%] among CN amyloid-negative participants; and from 16.4% [10.8%-22.1%] to 20.5% [18.8%-22.2%] among CN amyloid-positive participants. Among amyloid-negative participants with MCI and dementia, from age 60 to 80 years, the estimated prevalence of tau-PET-positivity increased from 3.5% [1.6%-5.3%] to 11.8% [7.1%-16.5%] and from 12.6% [4.5%-20.7%] to 15.9% [6.7%-25.1%] respectively. In contrast, among amyloid-positive participants with MCI and dementia, from age 60 to 80 years, the estimated prevalence of tau-PET-positivity decreased from 66.5% [57.0%-76.0%] to 48.3% [42.9%-53.8%] and from 92.3% [88.7%-95.9%] to 73.4% [67.5%-79.3%] respectively. APOE-e4 status primarily modulated the association of age with tau-PET-positivity estimates among CN and MCI amyloid-positive participants (Figure-2).

Conclusion: This large-scale multi-cohort study provides robust prevalence estimates of tau-PET-positivity, which can aid the interpretation of tau-PET in the clinic and inform clinical trial designs.